Complete tumor resection reverses neutrophilia-associated suppression of systemic anti-tumor immunity

Abstract

Objectives:

Tumor infiltrating neutrophils suppress T cell function, but whether neutrophils in circulation contribute to systemic immunosuppression is unclear. We aimed to study whether peripheral neutrophils that accumulate with tumor progression contribute to systemic immunosuppression, and if observed suppression of systemic anti-tumor immunity could be reversed with complete surgical tumor removal.

Materials and Methods:

Syngeneic murine oral cancers were established in immunocompetent mice. Proteomic and functional immune assays were used to study plasma cytokine concentration, peripheral immune frequencies, and systemic anti-tumor immunity with and without complete primary tumor resection.

Results:

Ly6G⁺ neutrophilic cells, but not other myeloid cell types, accumulated in the periphery of mice with progressing tumors. This accumulation positively associated with plasma G-CSF concentration. Circulating neutrophils were functionally immunosuppressive. Complete surgical tumor removal reversed the observed neutrophilia, with neutrophil frequencies returning to baseline in 21 days. Multiple independent functional assays revealed enhanced systemic anti-tumor immunity in mice following tumor resection compared to tumor-bearing mice, and the observed enhanced systemic immunity could be reproduced with selective neutrophil depletion.

Conclusions:

Complete primary tumor resection can reverse neutrophilia that develops during tumor progression and result in enhanced systemic anti-tumor immunity. Primary tumor removal relieves neutrophil-driven systemic immunosuppression and may itself contribute to the clinical benefit observed with neoadjuvant immunotherapy.

Introduction

At presentation, the majority of a head and neck squamous cell carcinoma (HNSCC) patient’s tumor-specific T cells are sequestered into the tumor microenvironment due to the development of a tissue resident-like phenotype induced by chronic T cell receptor (TCR) stimulation in the presence of TGF-β[1, 2]. Neoadjuvant immune checkpoint blockade (ICB)-based immunotherapy, administered prior to primary tumor resection, results in egress of tumor-specific T cells from the primary tumor and possibly the tumor-draining lymph nodes into circulation[1, 2]. Although more studies are needed to clearly associate this enhanced systemic anti-tumor immunity with clinical outcomes, this therapeutic effect resulting in enhanced systemic anti-tumor immunity may underlie the enhanced recurrence free survival (RFS) observed in numerous phase II clinical studies of neoadjuvant ICB in HNSCC not associated with HPV[3–6]. The addition of ICB to concurrent chemoradiation (CRT) in patients with newly diagnosed, unresectable HNSCC does not appear to improve RFS in initial studies[7, 8], but determining whether enhanced RFS is also observed following induction ICB administered prior to CRT requires additional clinical study.

Understanding the possible contributions of complete surgical removal of the primary tumor itself to the improved RFS observed with neoadjuvant immunotherapy may give insight into optimal treatment strategies for patients with newly diagnosed HNSCC unrelated to HPV. Through the release of cytokines, primary tumors induce emergency myelopoiesis, resulting in the significant expansion of myeloid cells in the bone marrow and circulation[9–11]. Myeloid cells, particularly neutrophilic cells, subsequently traffic through chemokine gradients into and accumulate in tumors[12, 13]. Neutrophilic cells in tumors are immunosuppressive, and their concentration in tumors is associated with response to ICB[6, 14]. Increased neutrophilia and neutrophilto-lymphocyte ratios in the peripheral blood of patients with cancer also inversely associates with poor prognosis and lack of response to ICB[15]. However, whether peripheral neutrophils are independently immunosuppressive and contribute to systemic immunosuppression in the context of a progressing primary malignancy remains unclear. Additionally, although initial studies have indicated that primary tumor removal can reverse neutrophilia, definitive experimental evidence demonstrating whether tumor removal can reverse neutrophil-associated systemic immunosuppression is needed.

Here, we utilized a syngeneic model of carcinogen-induced oral cavity SCC unrelated to HPV to study the development of tumor-associated systemic neutrophilia and immunosuppression, and the associated effects of primary tumor surgical resection on systemic anti-tumor immunity. Our data clearly demonstrate that tumors induce systemic immunosuppression mediated by neutrophils and that this can be reversed with complete primary tumor resection, supporting the possibility that complete primary tumor resection itself may, along with egress of tumor-specific T cells from the primary tumor into the periphery, contribute to the improved outcomes observed in patients treated with neoadjuvant immunotherapy and surgery.

Material and Methods

Cells and mice

Mouse oral cancer-1 (MOC1) cells are syngeneic on a C57BL/6 murine background and allow the study intact immunity in the host and tumor microenvironment. Original stocks of genomically characterized, mycoplasma free, low passage MOC1 cells were maintained in culture as described[16] and expanded for in vivo studies as needed. All in vivo experiments were performed in wild-type (WT) C57BL/6 (B6) mice or B6.129S7-Rag1^tm1Mom/J (RAG −/−) mice under an Animal Care and Use Committee–approved protocol. Tumors were established by flank subcutaneous injection of tumor cells (5 × 10⁶) in Matrigel (30% by volume). Peripheral blood was harvested via cardiac puncture into a heparinized 1 mL syringe from anesthetized mice prior to euthanasia. Bone marrow was isolated from femurs of euthanized mice using standard techniques. All complete primary tumor surgical resections were performed by the senior author (C. Allen) on anesthetized mice under sterile conditions as previously described[17]. Primary tumors but not tumor-draining lymph nodes (TDLN) were surgically excised at the time of tumor resection. Surgical wounds were closed with standard techniques using absorbable suture. Mice were administered meloxicam for analgesia after surgery for 3 days and monitored closely for changes in health in the postoperative setting. For all tumor cell challenge experiments, MOC1 tumor cells (5 × 10⁶) in Matrigel were injected subcutaneously in the contralateral flank. Tumor bearing mice were assessed for changes in tumor growth at least twice weekly, and tumor volume was calculated as: (length² × width)/2. Anti-mouse CD8 mAb (YTS 169.4, InVivoPlus), anti-mouse CD4 mAb (GK1.5), anti-mouse Ly6G (1A8) or isotype control mAb (LTF-2) was purchased from BioXCell and administered via intraperitoneal injection at 200 μg/injection in a total volume of 200 μL.

Cytokine ELISA

Whole blood was centrifuged at 1500xg for 10 minutes to obtain plasma. Plasma G-CSF, GM-CSF and M-CSF concentrations or supernatant IFNγ concentrations were determined using commercially available Quantikine ELISA kits from R&D Systems per manufacturer recommendations. % inhibition of IFNγ concentration was calculated as: [1−(IFNγ concentration in co-culture condition/ IFNγ concentration without co-cultured cells)] × 100.

Flow cytometry

Bone marrow leukocytes were isolated from marrow single cell suspensions generated by multiple passes through a 25-gauge syringe and RBC lysis (Biolegend). Peripheral blood leukocytes were isolated after centrifugation of whole blood and RBC lysis. Spleens and were crushed between frosted slides, RBC lysed, and filtered (70 μM). Tumors were minced, digested using a mouse tumor dissociation kit (Miltenyi) per manufacturer protocol and filtered (40 μM). Suspensions were washed with 1% BSA in PBS prior to blocking nonspecific staining with anti-CD16/32 (Biolegend) antibody. Single cell suspensions were stained with primary antibodies. Fluorophore-conjugated primary antibodies included anti-mouse CD45.2 clone 104, CD11b clone M1/70, Ly-6C clone HK1.4, Ly-6G clone 1A8, CD8 clone 53–6.7, NK1.1 clone PK136, CD4 clone GK1.5, FoxP3 clone FJK-16s, G-CSF clone 67604, F4/80 clone BM8, CD19 clone 6D5 and CD25 clone PC61.5.3. Cells were stained with antibodies for one hour, washed, and analyzed by flow cytometry on a BD Canto using BD FACS Diva software. All cells stained for cell surface marker analysis were stained with 7AAD to determine viability, and isotype controls and a “fluorescence minus one” method were used to determine staining specificity. FoxP3⁺ regulatory CD4⁺ T-lymphocytes (T_regs) were stained using the Mouse Regulatory T Cell Staining Kit #1 (eBioscience) per manufacturer protocol. Staining for other intracellular targets was performed with the Fixation and Permeabilization Buffer Set (eBioscience) per manufacturer protocol. Data were acquired on a FACSCanto using FACSDiva software (BD Biosciences) and analyzed on FlowJo software vX10.0.7r2[18].

Ly6G⁺ cell immunosuppression assay

Performed as described previously[19]. Briefly, total T cells were isolated from naïve B6 spleens using an autoMACS Pro Separator (Miltenyi Biotec) using the Pan T-Cell Kit (Miltenyi Biotec, negative selection) per the manufacturer’s protocol. T cells were stained with 5 μmol/L carboxyfluorescein succinimidyl ester (CFSE; Sigma) and stimulated using plate-bound CD3 (clone 145–2C11, eBioscience) and CD28 (clone 37.51, eBioscience) antibodies. Ly6G+ cells were similarly isolated from spleens of experimental or control mice using the Anti-Ly6G Microbead Kit (Miltenyi Biotec, positive selection). Ly6G+ cells were obtained from MOC1 tumor by processing tumor into single-cell suspensions by mincing, chemical (Murine Tumor Dissociation Kit, Miltenyi Biotec) and mechanical (gentleMACS, Miltenyi Biotec) dissociation per the manufacturer’s protocol, enrichment for total leukocytes using a 40/80% isotonic Percoll (Sigma) gradient (centrifuged at 325 × g for 23 minutes at room temperature), and selection using the anti-Ly6G Microbead Kit. Where indicated, CSFE-labelled T cells were cocultured with total unsorted splenocytes from WT B6 mice, or control or experimental sorted Ly6G+ cells from spleen or tumor at a 1:1 ratio for 4 hours prior to CD3/28 stimulation. Flow cytometry was used to quantify CFSE dilution after 72 hours of co-incubation. Proliferation was quantified as the average number of divisions for live CD8+ cells in the culture (division index) using FlowJo software [40]. % inhibition of T cell proliferation was calculated as: [1-(proliferation index in co-culture condition/proliferation index without co-cultured cells)] × 100. Media for all functional immune assays consisted of RPMI1640 supplemented with 10% FCS, 2 μmol/L β-ME, HEPES, nonessential amino acids, glutamine, and pen/strep antibiotics.

In vivo cytotoxicity assay

Whole splenocytes were harvested from naïve B6 mice and labeled with 2.5 ( high) or 0.25 μmol/L (low) CFSE. High CFSE-labeled cells were pulsed with 10 μg/mL of p15E_604–611 (KSPWFTTL) and CFSE-low cells were pulsed with control OVA_257–264 (SIINFEKL) peptide for 1 hour. Cells were washed, mixed at a 1:1 ratio, and adoptively transferred (2 × 10⁷ cells/mouse) into naïve, tumor bearing or tumor resected mice. Six hours later, splenocytes from recipient mice were harvested, and flow cytometry was used to determine the ratio of CSFE-high to CSFE-low cells. Antigen-specific cell killing was determined as: 1 − (r_naïve/r_{tumor bearing or tumor resected}) × 100, where r = (% CFSE-low cells)/(% CFSE-high cells).

Statistical analyses

All statistical analyses were performed using GraphPad Prism software. Significance between two parametric sets of data was determined with a t-test. Significance was determined between multiple sets of parametric data using a multiple comparisons test in a one-way analysis of variance (ANOVA). Significance between plasma G-CSF and cell concentrations in the blood and spleens of tumor bearing or tumor resected mice was determined using multiple unpaired t-tests with each timepoint considered independently. Significance between tumor growth curves was determined in a two-way ANOVA considering treatment group means for each timepoint. Significance between survival curves was determined with a log-rank test. A P-value of < 0.05 was considered significant.

Data availability

All processed data for this project is displayed in the main figures.

Results

Mice bearing progressing MOC1 tumors accumulate immunosuppressive Ly6G⁺ neutrophilic cells in the periphery

The mouse oral cancer-1 (MOC1) cell line is a carcinogen-induced squamous cell carcinoma cell line that results in progressive tumors after implantation into WT, immunocompetent mice (Fig. 1A). MOC1 tumors grow more rapidly in RAG^−/− mice, indicating that T cell immunity restrains the growth of MOC1 tumors in WT mice with fully functional immune systems. Several factors likely contribute to immune escape in MOC1 tumors that progressively grow in WT mice, including the accumulation of neutrophilic cells in the tumor microenvironment (Fig. 1B), previously shown to harbor T cell immunosuppressive capacity [12, 16–18]. Although previous work characterized immune profile changes in many different cell types in the periphery and tumor microenvironment in mice bearing progressing MOC1 tumors, notable in particular was the increased frequency of neutrophilic cells in the periphery [20], similar to observations of neutrophilia in patients with advanced head and neck cancer[21, 22].

Figure 1 – Progressing MOC1 tumors are restrained by T cell immunity but develop a tumor microenvironment enriched for neutrophilic cells.

A, a connected line dot plot shows tumor volume growth curves (n=10 mice per group) for wild-type (WT) or RAG^−/− mice engrafted with MOC1. Significance determined by 2-way ANOVA with Tukey’s multiple comparisons test.

B, stacked bar plots shows the distribution of different immune cell types within the total immune compartment in progressing MOC1 tumors (n=5 mice per timepoint).

To further study the accumulation of neutrophilic (Ly6G⁺) and monocytic (Ly6C⁺) myeloid cells in the periphery of mice bearing progressing MOC1 tumors, the fraction of Ly6G⁺ or Ly6C⁺ myeloid cells within the total immune compartment was determined by flow cytometry in the bone marrow (Fig. 2A), peripheral blood (Fig. 2B) and spleens (Fig. 2C) of wild type mice prior to and 10, 20, 28 and 38 days following MOC1 tumor implantation. The fraction of Ly6G⁺ cells, but not Ly6C⁺ cells, within the total immune compartment accumulated over time in the bone marrow, peripheral blood and spleen. To study possible associations between Ly6G+ cell accumulation and cytokines previously demonstrated to be important in myelopoiesis, plasma concentration of G-CSF, GM-CSF and M-CSF were determined in the plasma of tumor-bearing mice (Fig. 2D). Increased concentration of G-CSF, but not GM-CSF or M-CSF, was observed over time in tumor-bearing mice. Intracellular flow cytometry analysis of day 35 MOC1 tumors revealed that G-CSF is produced primarily by macrophages within the tumor microenvironment (Fig. 2E). This did not appear to be a tumor-specific finding as macrophages in the periphery (spleen) also appeared to produce G-CSF. These data indicated that accumulation of plasma G-CSF of macrophage origin associated with accumulation of neutrophilic cells in the bone marrow and periphery of mice bearing progressing MOC1 tumors, inferring a mechanistic link between tumor-derived G-CSF and the observed neutrophilia.

Figure 2 – G-CSF and immunosuppressive neutrophilic cells accumulate in the periphery of mice with progressing MOC1 tumors.

Representative flow cytometry dot plots of (A) bone marrow, (B) peripheral blood and (C) spleen samples (n=5 mice per timepoint) from non-tumor bearing mice (top dot plots) and day 38 tumor bearing mice (bottom dot plots) and connected line dot plots show the quantification of Ly6G+ or Ly6C+ myeloid cells as a percentage of live total leukocytes in tumor bearing mice during tumor progression. The concentration of cells in mice without tumors is shown as a dotted line in each plot. ***, P<0.001; significance was determined using multiple unpaired t-tests with each timepoint considered independently.

D, connected-line dot plots show the concentration of plasma growth factors in tumor bearing mice (n=5 mice per timepoint) during tumor progression. The plasma concentration of each factor in mice without tumors is shown as a dotted line in each plot.

E, representative flow cytometry dot plots (left) and dot plots showing quantification (right) of the percentage of CD31⁺CD45⁻ endothelial cells, CD31⁻CD45⁻ tumor cells, T cells, Ly6C⁺F4/80⁻ monocytes, Ly6C⁺F4/80⁺ macrophages, and Ly6G⁺ neutrophilic cells from the spleen or tumor of day 35 tumor-bearing mice (n=5) positive for G-CSF production.

F, representative CFSE-labelled T cell proliferation histograms and dot plots showing the quantification of the ability of Ly6G+ cells from the spleens or tumors of day 38 tumor bearing mice (TBM) or splenic Ly6G+ cells from non-tumor bearing mice or naïve total splenocytes to suppress T cell proliferation or IFNγ production. Significance determined with ANOVA multiple comparisons test.

Using an ex vivo T cell suppression assay, we next asked whether peripheral Ly6G⁺ cells are immunosuppressive, and if so, whether a difference in immunosuppressive capacity is observed between splenic Ly6G⁺ cells in mice with or without tumors (Fig. 2F). Ly6G⁺ cells sorted from the tumor microenvironment, known to be immunosuppressive[12, 20, 23, 24], were used as a positive assay control. Ly6G⁺ cells isolated from the spleens of day 38 tumor-bearing mice inhibited the ability to stimulated T cells to proliferate and produce IFNγ, but to a significantly lesser degree compared to Ly6G⁺ cells from the tumor. Peripheral Ly6G⁺ cells from tumor-bearing mice inhibited T cell function to a similar or modestly greater degree compared to peripheral Ly6G⁺ cells isolated from mice without tumors. Together, these data experimentally demonstrated that neutrophilic cells that accumulate in the periphery of MOC1 tumor bearing mice harbor the capacity to suppress T cell function and may contribute to systemic immunosuppression.

Primary tumor resection results in reversal of peripheral neutrophilia

Given the association between tumor progression and accumulation of plasma G-CSF, we next studied whether complete surgical removal of MOC1 tumors altered plasma G-CSF concentrations and accumulation of Ly6G⁺ cells in the periphery over time (Fig. 3A). Following complete MOC1 tumor surgical removal, plasma G-CSF (Fig. 3B), splenic Ly6G⁺ cells (Fig. 3C) and peripheral blood Ly6G⁺ cells (Fig. 3D) all reduced to levels similar those observed in mice without tumors in about 21 days. Although the exact kinetics of the reduction in G-CSF and Ly6G⁺ cells slightly varied from each other, significant reductions in each were observed within 7 days of tumor resection. These results indicated that the neutrophilia observed in mice with progressing MOC1 tumors can be reversed following complete primary tumor resection.

Figure 3 – Peripheral G-CSF and neutrophilic cells return to baseline after primary tumor resection.

A, illustration of the experiment designed to measure differences in the blood and spleens of tumor bearing or tumor resected mice.

B, a connected-line dot plot shows plasma G-CSF concentrations in tumor bearing or tumor resected mice. The black arrow indicates the time of primary tumor resection (day 20). The G-CSF concentration in mice without tumors is shown as a dotted line.

Connected-line dot plots show the quantification of Ly6G+ cells as a percentage of live total leukocytes in the spleens (C) or blood (D) of tumor bearing or tumor resected mice (n=10 mice total from two independent experiments for each condition and each timepoint). The black arrow indicates the time of primary tumor resection (day 20). The percentage of Ly6G+ cells in mice without tumors is shown as a dotted line in each plot.

For B, C and D, *, P<0.05; **, P<0.01, ***, P<0.001; significance was determined using multiple unpaired t-tests with each timepoint considered independently.

Primary tumor resection results in enhanced anti-tumor immunity

We next hypothesized that the observed reduction in peripheral immunosuppressive Ly6G⁺ cells after surgery would be associated with enhanced systemic anti-tumor immunity. To study this, we first performed an in vivo antigen-specific cytotoxicity assay that measures the ability of the immune system to selectively kill adoptively transferred cells loaded with the p15E_604–611 peptide (KSPWFTTL), a known H2-K^b-restricted MOC1 tumor-associated antigen (Fig. 4A). This assay revealed modest killing of p15E-pulsed transferred cells in mice bearing established day 34 MOC1 tumors that was significantly increased in mice 14 days after surgical tumor resection (Fig 4B), indicating enhancement of p15E-specific immunity after primary tumor removal. Notably, no evidence of p15E-specific cytotoxicity was observed in naïve, non-tumor bearing mice, indicating that the development of systemic antigen-specific anti-tumor immunity is dependent upon the presence of and may arise from a primary tumor.

Figure 4 – Enhanced systemic anti-tumor immunity is observed after primary tumor resection.

A, illustration of the in vivo cytotoxicity experiment that explores selective killing of adoptively transferred cells pulsed with p15E, a MOC1 tumor antigen.

B, dot plots show the ratio of CFSE low to CFSE high cells and the percentage of p15E-specific killing in naïve, MOC1 tumor bearing or tumor resected (day 14 after surgery) recipient mice (n=5 mice for the naïve condition and n=7 mice total from two independent experiments for tumor bearing or tumor resected conditions). n/s, not significant; significance determined with ANOVA multiple comparisons test.

C, a connected line dot plot shows the growth of challenge tumors in tumor bearing or tumor resected mice (day 14 after surgery) with or without CD8 or CD4 cell depletion (or isotype control, n=10 mice total from two independent experiments per condition) beginning one day prior to tumor challenge (depletion continued twice weekly for 4 weeks). The vertical dotted line corresponds to day 35 after challenge tumor injection. Challenge tumors in tumor bearing mice were measured to day 35 only, since this is the point at which the primary tumors in these mice begin to reach end-point criteria, requiring euthanasia. Significance determined by 2-way ANOVA with Tukey’s multiple comparisons test.

D, a dot plot shows the tumor volumes 35 days after challenge tumor injection. Significance determined with ANOVA multiple comparisons test.

E, a Kaplan-Meier survival curve shows survival in tumor resected mice with or without CD8 or CD4 depletion. Significance was determined with a log-rank test.

F, pot plots show splenic Ly6G⁺ cells 24 hours after administration of a single dose of an isotype control antibody before (top) or the 1A8 antibody (bottom). Inset percentages show the CD11b⁺Ly6G⁺ fraction of all live splenic immune cells.

G, a connected line dot plot shows the growth of MOC1 primary tumors with or without Ly6G depletion (starting at day 19, continuing twice weekly for 2 weeks). Significance determined by 2-way ANOVA with Tukey’s multiple comparisons test.

H, a connected line dot plot shows the growth of MOC1 challenge tumors engrafted at day 20 into the contralateral flank of mice bearing primary MOC1 tumors treated with a Ly6G depletion or isotype control at day 10 (one day before challenge tumor engraftment). Y-axis shows days after challenge tumor implantation. Significance determined by 2-way ANOVA with Tukey’s multiple comparisons test.

To further study hypothesized changes in systemic anti-tumor immunity following tumor resection, we performed tumor challenge experiments to directly measure systemic anti-tumor immunity. Here, MOC1 tumor cells are implanted in the contralateral flank forming secondary challenge tumors that can be followed for rate of engraftment and growth rate in mice bearing established primary MOC1 tumors for 14 days following primary tumor resection. Compared to mice bearing established primary tumors, that can only be followed to day 35 after injection of the secondary challenge tumor due to the primary tumor reaching end-point criteria, growth of secondary challenge tumors was significantly reduced in mice following primary tumor resection (Fig. 4C&D). To assess if the observed growth delay of secondary challenge tumors was T cell dependent, similar experiments were performed with T cell depletion. Depletion of CD8⁺ T cells, but not CD4⁺ T cells, eliminated the observed reduction in challenge tumor growth and survival benefit (Fig. 4E) in tumor-resected mice, indicating that the observed systemic anti-tumor immune response is CD8⁺ T cell dependent. Cumulatively, these data demonstrated that primary MOC1 tumor resection and reversal of the associated neutrophilia results in enhanced CD8-dependent systemic anti-tumor immunity of sufficient magnitude to delay the growth of challenge tumors.

It is possible that the delayed challenge tumor growth observed in tumor-resected mice could be due to the activity of memory anti-tumor T cells that develop in mice after tumor resection that are not present in tumor-bearing mice. To determine if elimination of Ly6G⁺ cells alone can enhance systemic anti-tumor immunity, we tested whether direct depletion of Ly6G⁺ cells would result in growth delay of primary or secondary challenge tumors in mice that have not undergone tumor resection. Near complete depletion of Ly6G⁺ cells can be achieved after single 1A8 depleting antibody injection (Fig. 4F), and previously published data indicates that Ly6G⁺ cells are durably depleted for at least 6 days[20]. Depletion of Ly6G⁺ cells in established MOC1 tumors, starting at day 19, did not delay primary tumor growth (Fig. 4G). This is consistent with prior observations that inhibition of neutrophils in established lesions does not alter tumor progression[12]. However, depletion of Ly6G+ cells beginning one day before challenge tumor engraftment both reduced the engraftment rate compared to control (70% vs 100%) as well as significantly delayed the secondary challenge tumor growth (Fig. 4H). These data suggest that although depletion of neutrophilic cells from established primary tumors does not result in enhanced immunity of sufficient magnitude to alter tumor progression, depletion of peripheral neutrophilic cells does enhance systemic immunity in a magnitude sufficient to reduce the engraftment and growth of secondary challenge tumors, validating the functional importance of neutrophilic cells in contributing to tumor-induced systemic immunosuppression.

Discussion

Our results confirm prior reports[25–27] and demonstrate that a progressing primary tumor in an immunocompetent host can drive myelopoiesis that results in systemic suppression of anti-tumor immunity that is due at least to neutrophilia. Additionally, we show with multiple independent assays that surgical removal of the primary tumor allows normalization of peripheral neutrophil counts, resulting in enhancement of a systemic, tumor-specific immune response that formed upon immune system exposure to the primary tumor. These results suggest that partial or complete reversal of systemic immune suppression following complete primary tumor removal may itself contribute to the enhanced systemic anti-tumor immunity observed following neoadjuvant immunotherapy. Since neoadjuvant immunotherapy results in egress of tumor-specific T cells into circulation in patients with newly diagnosed HNSCC[1, 2], subsequent primary tumor resection may partially or completely reverse neutrophilinduced systemic immune suppression and allow better systemic immune function compared to immune function if the primary tumor were left in place. Specifically, our studies demonstrate that the magnitude of systemic anti-tumor immunity is greater when the primary tumor is surgically removed compared to when it is left in place at a single time point. Determining whether primary tumor removal results in greater duration of systemic anti-tumor immunity would require additional study.

Our study does not exclude that similar observations can be made after non-surgical treatment, such as chemoradiation. Although radiation may play a role in activating local anti-tumor immunity through multiple distinct mechanisms, reviewed in[28], radiation also results in systemic lymphopenia that may inhibit systemic anti-tumor immunity[29]. Similarly, chemotherapy also results in lymphopenia but may activate immunity locally depending on dose and schedule (reviewed in [30]). However, to the best of our knowledge, reversal of neutrophilia-associated systemic immunosuppression following complete tumor resolution with chemotherapy and/or radiation treatment has not been reported. If reduced neutrophil counts are observed following chemotherapy or radiation, attributing this observation to treatment of the primary tumor and biologic reversal of tumor-driven neutrophilia or simply to direct treatment-related neutropenia may be difficult. Ultimately similar experiments to those reported here that directly measure tumor-derived G-CSF and systemic anti-tumor immunity with functional studies would be needed to determine if similar observations can be made following chemotherapy and/or radiation treatment in models where such treatment can result in complete tumor resolution. MOC1 tumors cannot be cured with cisplatin-based chemotherapy or radiation treatment[31, 32].

The MOC1 model was chosen for this study due to its fidelity to human HNSCC[33], its ability to result in systemic anti-tumor immunity[34], the known accumulation of neutrophilic cells in the TME and periphery in MOC1 tumor-bearing mice[20], and its ability to be completely surgically resected[17]. A disadvantage of the MOC1 model is that it is unknown whether establishment of a MOC1 primary tumor in immunocompetent mice results in disseminated tumor cells (DTCs). In this study, we used direct measures of anti-tumor immunity to study the effect of primary tumor resection, but another important study would be to determine whether tumor resection and resolution of neutrophilia resulted in elimination of DTCs. Others have demonstrated that primary tumor resection can result in the elimination of DTCs, but this effect appears to be model-dependent[35].

Another important consideration is whether systemic anti-tumor immunity is differently altered after primary tumor resection when obvious distant disease is already established. We were unable to study this in our model since it does not spontaneously develop distant metastasis. Observations of relatively increased growth kinetics of metastatic disease following primary tumor resection have been made. Krall et al. observed that surgical wounding induces a systemic inflammatory response that transiently abrogates tumor-specific T cell immunity, leading to growth of a distant tumor[36]. However, multiple other studies indicate a reversal of tumor-induced immunosuppression after surgical resection, even in the presence of significant metastatic disease[26, 35, 37, 38]. Whether primary tumor resection leads to enhanced distant tumor growth may again be model dependent and relate to specific details of the experimental approach used.

These findings should be considered with the limitations of our study. Owing to the work-intensive nature of surgical resection in mice, we performed the described studies using only one syngeneic model. The use of more than one model could allow inference of generalizability of these findings. The one model we selected to use does not spontaneously result in distant metastasis and it is unknown whether it results in the presence of DTCs. Similar studies performed in a model that has either of these features could allow additional study of the ability of primary tumor resection to abrogate or alter distant metastasis formation or the presence of DTCs. Additionally, although not the primary focus or finding of this work, we did not thoroughly study our finding that the origin of G-CSF that accumulates in the plasma of mice bearing progressing tumors appears to be macrophages. To do this properly would require the study of G-CSF production at different timepoint in tumor progression as well as whether G-CSF production by circulating macrophages is reduced following tumor resection. Such studies could provide valuable insight into the interplay between tumorigenesis, tumor progression, and the development of G-CFS-associated neutrophilia.

In conclusion, our work demonstrates that enhanced systemic anti-tumor immunity is observed following primary tumor resection compared to when the tumor remains present. Given that neoadjuvant immunotherapy results in the increased detection of tumor-specific T cells in the periphery in patients with HNSCC, formation of a more immune permissive systemic environment following primary tumor resection may play an important role in preventing disease relapse. Together, these data support the continued clinical study of immunotherapy prior to complete primary tumor surgical resection in efforts to improve RFS for patients with HNSCC and suggest that the surgery itself may be an important aspect of this treatment strategy.

Highlights:

• Immunosuppressive neutrophilia develops with tumor progression and drives systemic immunosuppression

• Complete primary tumor removal reverses neutrophilia

• Enhanced tumor-specific systemic immunity is observed after tumor removal

• Primary tumor removal itself may contribute to the observed efficacy of neoadjuvant immunotherapy