Lung Ablation with Irreversible Electroporation Promotes Immune Cell Infiltration by Sparing Extracellular Matrix Proteins and Vasculature: Implications for Immunotherapy

Abstract

Background: This study investigated the sparing of the extracellular matrix (ECM) and blood vessels at the site of lung irreversible electroporation (IRE), and its impact on postablation T cell and macrophage populations.

Materials and Methods: Normal swine (n = 8) lung was treated with either IRE or microwave ablation (MWA), followed by sacrifice at 2 and 28 days (four animals/timepoint) after treatment. En bloc samples of ablated lung were stained for blood vessels (CD31), ECM proteins (Collagen, Heparan sulfate, and Decorin), T cells (CD3), and macrophages (Iba1). Stained slides were analyzed with an image processing software (ImageJ) to count the number of positive staining cells or the percentage area of tissue staining for ECM markers, and the statistical difference was evaluated with Student's t-test.

Results: Approximately 50% of the blood vessels and collagen typically seen in healthy lung were evident in IRE treated samples at Day 2, with complete destruction within MWA treated lung. These levels increased threefold by Day 28, indicative of post-IRE tissue remodeling and regeneration. Decorin and Heparan sulfate levels were reduced, and it remained so through the duration of observation. Concurrently, numbers of CD3⁺ T cells and macrophages were not different from healthy lung at Day 2 after IRE, subsequently increasing by 2.5 and 1.5-fold by Day 28. Similar findings were restricted to the peripheral inflammatory rim of MWA samples, wherein the central necrotic regions remained acellular through Day 28.

Conclusion: Acute preservation of blood vessels and major ECM components was observed in IRE treated lung at acute time points, and it was associated with the increased infiltration and presence of T cells and macrophages, features that were spatially restricted in MWA treated lung.

Introduction

Percutaneous image guided ablation has gained broad application for the treatment of small primary and metastatic tumors in the lung.^1–10 Several ablation techniques have been used in patients, including radiofrequency ablation (RFA),^3–5 microwave ablation (MWA),^6,7 cryoablation,^8,9 and irreversible electroporation (IRE).¹⁰ These techniques are broadly classified into thermal and nonthermal ablation based on their working principle, wherein the former utilize prolonged alteration of tissue temperature for tumor destruction.^11,12

Cell death due to thermal or nonthermal ablation can both trigger an inflammatory or immune response,^13–15 but the two ablation modalities have divergent effects on key components of the tumor microenvironment that orchestrate immune cell infiltration. Thermal ablation techniques such as RFA and MWA cause coagulative damage to blood vessels,^16,17 resulting in rapid loss of perfusion within the tumor.^17,18 Sustained high temperatures (60°C or higher) that is necessary for efficacious thermal ablation also denatures the extracellular matrix (ECM) proteins within the treatment volume.^19,20 In contrast, nonthermal ablation techniques such as IRE have been reported to spare blood vessels and the ECM within the treatment volume despite transient increase in tissue temperature during energy delivery.^17,21–24

Functional blood vessels are essential for the recruitment of circulating immune cells and adequate tissue perfusion is necessary for T cell survival and function.^25–27 Likewise, ECM proteins such as Heparan sulfate and Decorin provide binding sites for chemokines that guide immune cell trafficking and collagen is important for cell migration.^24,28,29 Further, it is standard of care to extend the ablation zone into peri-tumoral healthy tissue to obtain a 5 mm or greater margin.^5,20,30–33 However, the impact of postablation immune cell infiltration in normal lung is largely unknown and not amenable for study in commonly used mouse or rodent models. While RFA and MWA represent the predominant thermal ablation modalities used in the lung, the latter has shown numerous strengths, gaining rapid clinical adoption.^34,35 Therefore, the objective of this study was to compare and quantify differences in the status of ECM proteins and vasculature after thermal (MWA) and nonthermal ablation (IRE) of normal porcine lung and determine its impact on immune cell infiltration.

Materials and Methods

Animal care and ablation protocol

All animal work was performed with approval from the Institutional Animal Care and Use Committee (Protocol # 08-08-016). Tissue samples from eight Yorkshire swine (weight: 50–70 kg, equal male, and female) were included for this study. The animals underwent image guided MWA or IRE under inhaled isoflurane anesthesia (2% Aerrane; Baxter, Deerfield, Illinois). CT (Lightspeed; GE Healthcare, Milwaukee, WI) or C-arm (Innova; GE Healthcare) image guidance was used to perform probe placement and monitor treatment delivery. All ablations were performed in unilateral lung using a single treatment modality.

MWA was performed by delivering energy at 100 W power for 2 min (17 kJ, 2450 MHz) using a single Emprint™ Ablation probe (Medtronic, Boulder, CO). IRE (Nanoknife; Angiodynamics, Latham, NY) was performed by placing two parallel needle electrodes (2 cm exposure, 1.5 cm spacing) and delivering ninety pulses (70 μs pulse length, delivered at ∼1 Hz frequency, with cardiac gating) while maintaining an electric field strength of ∼1500 V/cm. The energy settings were chosen to produce an equivalent sized ablation using both modalities (IRE: 1 .55 ± 0.28 cm vs. MWA: 1.5 ± 0.19 cm, p = 0.873). All ablations were performed after administering a paralytic (pancuronium or rocuronium, 0.1 mg/kg, Hospira; Lake Forest, IL) to facilitate breath holds during probe placement and reduce neuromuscular activation during pulse delivery for IRE. Two animals from each treatment group were sacrificed by 2 (early) or 28 (late) days after ablation using intravenous injections of pentobarbital sodium and phenytoin sodium. Complete procedure details, animal care protocols, and other information for these experiments have been previously reported in these studies.^18,22,32

Histopathologic evaluation

Gross dissection was performed, and lungs were harvested en-bloc. Lung ablations were identified per techniques defined by Kodama et al. for MWA and Deodhar et al. for IRE.^18,22 The ablated tissue and the surrounding normal lung were harvested, and fixed by immersion in 10% neutral buffered formalin, embedded in paraffin, sectioned at 5 μm thickness, and stained with hematoxylin/eosin (H&E) and Masson's Trichrome for collagenous ECM. The sectioned samples were also stained with antibodies for blood vessels (CD31, 1:100 dilution, ab28364; Abcam), ECM components of Heparan sulfate (1:100 dilution, ab2501; Abcam), Decorin (1:300 dilution, LSB8177; LS Bio), T cells (CD3; 1:100 dilution, ab16669; Abcam), and Macrophages (IBA1; 1:1000 dilution, ab5076; Abcam). Antibodies were validated for staining swine tissue with dose optimization and appropriate positive and negative control tissues.

Assessment and statistical analysis

Samples from a total of eight animals (one sample per animal) were included in this study, yielding two independent specimens per treatment modality and time point. Complete details on ablation size and treatment efficacy for each modality are described in detail by Kodama et al. for MWA, and by Deodhar et al. for IRE.^18,22 All stained slides were scanned at high resolution (40 × magnification) using a Leica AT2 automated scanner instrument, allowing complete visualization of the treated and the surrounding normal lung. H&E stained images were used to delineate three separate tissue regions for analysis, namely, normal lung, regions having inflammatory infiltrate with (Day 2) or without (Day 28) cell death, and central necrotic regions (CNR) (Fig. 1 and Fig. 2). Analysis and quantification were performed separately for the peripheral inflammatory region (IR) and CNR. From the scanned images, random locations were selected for acquiring 10 independent images from each region (IR and CNR), and for every specimen. Data acquired from both the samples for every treatment/time point were pooled for analysis.

FIG. 1.

Schematic illustrations of IR and CNR in IRE and MWA of normal swine lung. CNR, central necrotic regions; IR, inflammatory regions; IRE, irreversible electroporation; MWA, microwave ablation.

FIG. 2.

(A, B) Gross histology images show two regions in IRE treated lung (*, normal lung; **, IR). White arrow in Day 2 sample shows blood filled tract through which IRE applicator was inserted. Box and insert show high magnification of alveoli in the central IRE treated region of Day 2 sample. Alveolar space (white asterisk) is filled with red blood cells and inflammatory cells. A small number of cells on alveolar wall show pyknosis of nuclei (white arrowheads) and central necrotic zone was not remarkable. Scale bar in the insert shows 50 μm. (C, D) Gross histology images show three regions in MWA treated lung (*, normal lung; **, IR; ***, CNR). White arrow in Day 2 sample shows a hole in which MWA antenna was inserted. Box and insert show high magnification of alveoli in CNR of Day 2 sample. Cell necrosis was identified based on shrinkage and pyknosis of nuclei. Scale bar in the insert shows 100 μm.

Analysis and quantification of staining data were performed by two independent observers (M.F., G.S.). Discrepant interpretation was resolved via consensus review. The pictures were analyzed with an image processing software (ImageJ, NIH) to count the number of cells with positive staining in the field of view (FOV) (defined as an entire image taken at 40 × magnification) or measure the percentage area of tissue staining positive for ECM proteins. Mean and standard deviation (SD) calculation were performed on the histopathologic examination by using SAS software (SAS Institute, Cary, NC). Comparison was achieved using a Student's t-test when comparing between two time points in the same region. In addition, one-way ANOVA followed by post hoc Tukey test was performed when comparing each region. The differences were considered statistically significant at p-value of <0.05.

Results

IRE and MWA treated lung demonstrate distinct responses in the postablation setting.

H&E-stained images of IRE treated lung from the early time point revealed a widespread hemorrhagic zone with cell death and inflammatory cell infiltration, and a small CNR (Fig. 2A). In contrast, H&E staining of MWA treated lung revealed two separate regions, a large acellular CNR and a peripheral IR (Fig. 2B). By Day 28, the entire region of IRE treated lung was interpreted as an IR (Fig. 2C), while MWA treated specimens retained distinguishable zones CNR and IR (Fig. 2D). Relative change in ablation size for each treatment modality was not measured as the samples were derived from unrelated animals.

IRE treated lung exhibited greater number of blood vessels at both the acute and late time point

Representative results of CD31 immunohistochemistry are shown in Figure 3A. At the early time point, fewer blood vessels were present in IRE treated lung based on CD31 staining (0.30 ± 0.48 vessels/FOV, at Day 2) (mean ± SD) when compared to the surrounding normal lung (2.05 ± 0.76 vessels/FOV, p < 0.0001). By Day 28, considerable wound healing related neovascularization was evident, with considerably greater number of blood vessels when compared to normal lung (3.70 ± 1.7 vessels/FOV, p < 0.0001) (Fig. 3B). Neovasculature were also observed in the peripheral IR of MWA at Day 28, but not at Day 2 (0 ± 0 vs. 1.60 ± 1.78 vessels/FOV, p = 0.01) (Fig. 3C). The CNR in MWA treated lung was avascular and remained so at both time points, with no meaningful CD31 staining (Fig. 3D). Comparison of Day 2 IRE and MWA samples did not indicate a significant difference in the number of blood vessels (Fig. 3E). However, the number of CD31positive vessels in IR of IRE and MWA was significantly higher than that in CNR of MWA (p < 0.0001 in IRE and p < 0.05 in MWA) at Day 28 (Fig. 3F). Further, the number of CD31-positive vessels in the IR of IRE samples were greater when compared to the corresponding region in the MWA samples (p < 0.01).

FIG. 3.

Result of CD31 immunohistochemistry study. (A) Blood vessels within the IR were preserved, but the CNR of MWA was avascular. White arrows show CD31 positive stained vessels (brown). (B, C) On Day 2, the number of CD31 positive vessels was fewer in IR of both IRE and MWA. By Day 28, increased numbers of blood vessels were observed when compared to normal lung. (D) There was no CD31 positive vessels in CNR of MWA treated lung at both time points. (E) On Day 2, there was no significant difference between each region. (F) On Day 28, the number of CD31 positive vessels in IR of IRE and MWA was significantly higher than that in CNR of MWA. In addition, the number of CD31 positive vessels in IR of IRE was also significantly higher than that of MWA. Dashed bars in the graph show the mean number of CD31 positive stained vessels in untreated normal lung (*, ** and **** mean p < 0.05, p < 0.01, and p < 0.0001, respectively. Scale bar shows 50 μm). FOV, field of view.

Collagen and ECM proteins are spared in IRE treated lung

Analysis of Masson's Trichrome stained Day 2 samples (Fig. 4A) revealed a slight reduction of collagen content in IRE treated lung (6.7% ± 2.07%) in comparison to normal lung (10.2% ± 1.81%, p < 0.0001), but levels increased by Day 28 (27.6% ± 6.45%, p < 0.0001) (Fig. 4B), indicative of scar formation in the ablation zone. Similar to IRE samples, collagen levels were somewhat decreased in the IR of MWA samples at Day 2 (2.5% ± 0.93%), then recovering to higher levels by Day 28 (27.8% ± 6.29%, p < 0.0001) (Fig. 4C). In contrast, no positive staining was observed in the CNR of MWA treated lung at both time points (Fig. 4D). When comparing each region, more Masson's Trichrome positive area was observed in IR of both treatment modalities than the CNR of MWA on Day 2 (p < 0.0001 in IRE and p < 0.001 in MWA). IR of IRE treated lung showed more Masson's Trichrome positive area when compared to the IR of MWA samples (p < 0.0001) (Fig. 4E). On Day 28, higher Masson's Trichrome positive area was observed in IR of both treatment modalities when compared to the CNR of the MWA samples (p < 0.0001) (Fig. 4F).

FIG. 4.

Result of Masson's Trichrome staining for collagen. (A) Collagen (blue) was preserved after IRE and the percentage increased in IR because of new collagen deposit. On the other hand, collagen deposit occurred only in IR after MWA. There was no Masson's Trichrome positive staining in the CNR after MWA. (B, C) A slight reduction of collagen positive area was observed in IR of both IRE and MWA. On Day 28, collagen positive area in IR recovered to higher levels than untreated normal lung. (D) No positive staining was observed in CNR of MWA treated lung at both timepoints. (E) On Day 2, more Masson's Trichrome positive area was observed in IR of both modalities than the CNR of MWA. In addition, IR of IRE had larger region of positive staining than the IR of MWA. (F) On Day 28, more Masson's Trichrome positive area was observed in IR than CNR of MWA. Dashed bars in quantification results show mean percentage of Masson's Trichrome positive tissue area in untreated normal lung (*** and **** mean p < 0.001 and p < 0.0001. Scale bar shows 50 μm).

Representative results of Heparan sulfate and Decorin immunohistochemistry are shown in Figures 5A and 6A. After IRE, the percentage area of the specimen staining positive for Heparan sulfate (Day 2: 6.7% ± 5.20%, Day 28: 4.3% ± 1.25%) and Decorin (Day 2: 8.6% ± 2.03%, Day 28: 9.5% ± 2.03%) was reduced in comparison to the surrounding normal lung (Heparan sulfate; 12.1% ± 3.85%, Decorin; 22.0% ± 6.69%), but nevertheless sparing of these proteins was observed (Figs. 5B and 6B). IR of MWA samples presented positive staining for Heparan sulfate but not Decorin (Day 2: 5.8% ± 2.29%, Day 28: 5.3% ± 2.30%) and the region of positive staining was comparable to that of IRE (Day 2: p = 0.601, Day 28: p = 0.273) (Figs. 5C and 6C). In contrast, there was no positive staining for these ECM proteins in CNR of MWA at both the time points (Figs. 5D and 6D). Larger region of positive staining for Heparan sulfate was seen in the IR of both treatment modalities than the CNR of MWA samples at both Day 2 (p < 0.001 in IRE and p < 0.01 in MWA, Fig. 5E) and Day 28 (p < 0.0001, Fig. 5F). Larger region of positive staining for Decorin was observed in the IR of IRE samples when compared to the IR or CNR of MWA samples at Day 2 (p < 0.001, Fig. 6E) and Day 28 (p < 0.0001, Fig. 6F).

FIG. 5.

Result of Heparan sulfate immunohistochemistry (A) Heparan sulfate (brown staining) was preserved in IR after IRE and was restricted in IR of MWA but was absent in the CNR of MWA. (B, C) Compared to surrounding normal lung, Heparan sulfate positive area was reduced, but relatively preserved in IR after IRE and MWA. (D) At both the time points, there was no positive staining for Heparan sulfate in CNR after MWA. (E) On Day 2, more Heparan sulfate positive area was observed in IR of both modalities than the CNR of MWA. (F) On Day 28, findings mirrored that of the earlier time point. Dashed bars in quantification results show mean percentage of positive tissue are with Heparan sulfate in untreated normal lung (**, ***, and **** mean p < 0.01, p < 0.001, and p < 0.0001, respectively. Scale bar shows 50 μm).

FIG. 6.

Result of Decorin immunohistochemistry. (A) Decorin (brown staining) was preserved in IR after IRE. These was no Decorin positive stained tissue after MWA in both IR and CNR. (B) Compared to surrounding normal lung, Heparan sulfate positive area was reduced, but relatively preserved in IR after IRE. (C, D) No positive staining was observed in IR and CNR of MWA treated lung at both timepoints. (E) On Day 2, more Decorin positive area was observed in the IR of IRE than the IR and CNR of MWA. (F) On Day 28, findings mirrored that of the earlier time point. Dashed bars in quantification results show mean percentage of positive tissue are with Decorin in untreated normal lung (*** and **** mean p < 0.001 and p < 0.0001, respectively. Scale bar shows 50 μm). IR, inflammatory regions.

Increased infiltration of CD3 positive T cell and Iba1 positive macrophage was observed throughout IRE treated lung

Representative results of CD3 and Iba1 immunohistochemistry are shown in Figures 7A and 8A. The numbers of T cells (8.0 ± 2.71 cells/FOV) and macrophages (8.2 ± 2.74 cells/FOV) were lower in IRE treated lung at Day 2 in comparison to surrounding normal lung (T cells; 13.9 ± 8.24/FOV, macrophages; 10.2 ± 4.37 cells/FOV). The numbers of T cells (25.6 ± 7.79 cells/FOV) and macrophages (16.1 ± 4.75 cells/FOV) were greatly increased by Day 28 (Figs. 7B and 8B). Cell infiltration was restricted to the IR in MWA samples, while the numbers of T cells (Day 2: 6.9 ± 2.02 cells/FOV at, Day 28: 18.0 ± 7.07 cells/FOV, p = 0.0002) and macrophages (3.0 ± 1.63 cells/FOV at Day 2, 24.5 ± 8.07 cells/FOV at Day 28, p < 0.0001) increased between the two time points (Figs. 7C and 8C). In comparison, no T cells or macrophages were observed in the CNR of MWA treated lung (Figs. 7D and 8D). More T cells were present in the IR of both treatment modalities than CNR of MWA at Day 2 (p < 0.0001 in IRE and p < 0.001 in MWA, Fig. 7E) and Day 28 (p < 0.001 in IRE and p < 0.01 in MWA). In addition, more T cells were observed in IR of IRE samples than the corresponding region in MWA samples (p < 0.05, Fig. 7F). There were more macrophages in IR of samples from both the treatment modalities when compared to the than CNR of MWA samples at both Day 2 and 28. More macrophages were observed in IR of IRE samples at Day 2 but this observation was reversed at Day 28 where the IR of MWA had a greater number of macrophages (p < 0.01, Fig. 8E, F).

FIG. 7.

Result of CD3 immunohistochemistry study. (A) On Day 2, CD3 positive cells (brown) were observed in IR after IRE and MWA but not in CNR after MWA. On Day 28, CD3 positive stained T cells increased in IR after IRE and MWA. (B, C) T cells were less than normal lung tissue, but relatively preserved. The number of T cells was significantly higher on Day 28 than on Day 2 in IR of IRE and MWA. (D) At Day 2, only a few T cells were observed in CNR but, at Day 28, no T cells were observed in the CNR of MWA treated lung. (E) On Day 2, there were more T cells in the IR of both treatment modalities than the CNR of MWA. (F) On Day 28, greater number of T cells were observed in IR of IRE than the IR or CNR of MWA samples. Dashed bars in quantification results show the mean number of CD3 positive stained T cell (brown) in untreated normal lung (*, **, ***, and **** mean p < 0.05, p < 0.01, p < 0.001, and p < 0.0001, respectively. Scale bar shows 50 μm).

FIG. 8.

Result of Iba1 immunohistochemistry study. (A) Macrophages (brown) penetration was seen in IR after IRE and MWA on Day 2 while no positive cells were observed in CNR after MWA. On Day 28, macrophage populations increased in IR after IRE and MWA. (B, C) On Day 2, the number of macrophages was reduced but maintained comparing with normal lung tissue. On Day 28, macrophages population increased in IR after IRE and MWA, higher than normal lung tissue. (D) After MWA, there were no Iba1 stained macrophages in CNR. (E) On Day 2, there were more macrophages in the IR of both treatment modalities than the CNR of MWA. In addition, more macrophages were observed in the IR of IRE when compared to the IR of MWA. (F) On Day 28, more macrophages were observed in IR of MWA than the IR of IRE. Dashed bars in quantification results show the mean number of Iba1 positive stained macrophages in untreated normal lung (**, ***, and **** means p < 0.01, p < 0.001, and p < 0.0001, respectively. Scale bar shows 50 μm).

Discussion

During percutaneous ablation, establishing a treatment margin extending 5 mm or more into surrounding healthy tissue is considered essential for effective local disease control.^5,20,30–33 Our experiment was performed in normal swine lung as it allowed evaluation of this clinically relevant setting in an animal model that can recapitulate patient size and ablation volume. Furthermore, our approach precluded the confounding effects of tumor-secreted chemokines or cytokines that serve as attractants to immune cells even before intervention. Results of our study indicates that early sparing of vasculature and ECM proteins in IRE treated normal lung promotes rapid infiltration and subsequent proliferation of T cells and macrophages throughout the ablated tissue (Supplementary Fig. S1). In patients, this may translate to penetration of immune cells throughout the ablated tumor and its periphery. Such activity in MWA treated lung was restricted to the IR, which in the clinical setting may largely be comprised of otherwise healthy lung at the ablation margin.

As ablation of lung tumors is performed with curative intent, the modality used to treat a given patient is guided by its anticipated efficacy while minimizing potential adverse effects. Thermal ablation techniques have demonstrated excellent efficacy for the treatment of tumors smaller than 3 cm with an acceptable adverse event profile.^{3,6,7,36–39} Preliminary evaluation of IRE in the lung yielded unsatisfactory outcomes, dampening further utilization in the clinic.¹⁰ This might be due to heterogeneity in the bioelectric properties of tumors, and the unfavorable electrical conductive properties of the lung.^10,32,33 However, emerging evidence suggests that besides local tumor control, ablation can serve as an adjuvant for augmenting cancer immunotherapy. Cell death during ablation releases tumor-associated antigens and damage-associated molecular patterns, rapidly mobilizing innate immune cells to the treatment site.^19,40 Improved antigen presentation by macrophages and dendritic cells can stimulate T cells, increasing their infiltration into ablation-treated and distant untreated tumors.^40,41 Preclinical research has shown that IRE mediated immune activity can be leveraged for improving locoregional or systemic cancer control by combination therapy with immune checkpoint inhibitors (ICI).^15,42,43 These are of potential interest in patients with primary lung cancer where ICI is considered a first line of therapy for advanced stages. While IRE has proven unsatisfactory for the curative ablation of lung tumors, our findings lend support for its examination as a neo-adjuvant during cancer immunotherapy using ICI or other approaches.

Under favorable conditions, tumor ablation can result in an “abscopal effect,” which leads to remote tumor regression. This effect has been previously described in case reports after thermal ablation and in the preclincial setting after IRE.^15,44–46 In contrast, Ahmed et al., Kumar et al., and Rozenblum et al. have reported that wound healing after thermal ablation can have off-target tumorigenic effect in animal models, wherein factors related to tissue repair, inflammation, and angiogenesis can promote remote tumor growth.^47–49 Our findings indicate that ablation of normal lung attracts both T cells and macrophages, wherein their specific phenotype informs their function, and influences their anticancer activity. Macrophages exposed to necrotic cells secrete transforming growth factor-beta1 that can suppress CD8-T and natural killer cell activation and promotes regulatory T cell proliferation.^19,50–52 After thermal ablation, an immunosuppressive environment may therefore emerge in the periphery of the ablation, where denatured ECM and loss of vasculature further prevent the immune cells migration into central region of the ablation containing tumor antigens. In contrast, IRE allows both T cells and macrophages to penetrate throughout the ablated tumor, where combination with appropriate adjuvants may increase antitumor response by promoting T cell activity.^15,42 These are topics of broad interest that require further analysis in future experiments.

Our study provides promising evidence supporting the evaluation of IRE in lung tumors for noncurative ablation but is not to be considered definitive or conclusive. The small number of animals included in the study, limited number of time points, and lack of detailed analysis on immune cell phenotype and activity are all limitations of our work. Large animal studies are subject to stringent ethical care guidelines that limit the number of animals that can be included for studies when compared to the large cohort sizes that are feasible in rodent models. Likewise, while we used robust statistical techniques to account for our smaller sample size, the limited number of samples per animal is a source of bias that has to be investigated further in future studies. Our study was based on image analysis of IHC samples instead of flow cytometry, a more standard technique for immunoprofiling. However, our choice of using image analysis was a consequence of the large size of our tissue samples that creates difficulties while processing for evaluation with flow cytometry. Ablation is well known to trigger changes in circulating lymphocyte populations, but systemic response to treatment was not an aim of this study, and hence immune status of peripheral blood and its chemistry were not evaluated. While we did not perform phenotypic analysis of the immune cells at the site of ablation, other studies report that these populations typically skew to M1 polarized macrophages and tumor cognizant CD8⁺ and CD4⁺ T cells.^42,53–58 It would be of interest to replicate our findings in a tumor model as the ECM and vasculature in that setting are distinct from our current study using healthy lung. The consequence of preserving ECM and vasculature within tumors may have implications besides immune cell infiltration, as evidenced by prior studies where incomplete IRE of pancreatic tumors in mice was associated with tumor regrowth.⁵⁹

Conclusion

Treatment of healthy lung with IRE results in tissue ablation while largely sparing vasculature and ECM components, which are completely destroyed during MWA. This correlates with the early infiltration and increased presence of T cells and macrophages throughout IRE treated lung. Such findings are restricted to the peripheral regions of MWA samples, with no presence of immune cells in the CNR of the ablation.

Authorship Confirmation Statement

Experimental design: M.F., P.S.A., G.S. Data collection: M.F., E.U., Y.K., D.E.D., S.B.S. Analysis and interpretation: M.F., Y.K., G.S. Article preparation and editing: M.F., E.U., Y.K., D.E.D., P.S.A., S.B.S., G.S. All co-authors have reviewed and approved the article submitted for review. The work reported in this article has been submitted solely to this journal, and is not published, in press or submitted elsewhere.

Author Disclosure Statement

The authors report no relevant conflict of interest related to the work presented here. S.B.S. is a consultant to BTG, Johnson & Johnson, XACT, Adegro, and Medtronic. S.B.S. has funding support from GE Healthcare and Angiodynamics, and holds stock in Aperture Medical. G.S. has received consulting fees from Farapulse and Intuitive Surgical, and holds stock options in Aperture Medical. P.S.A. has received research funding from ATARA Biotherapeutics and Acea Biosciences, has served on the Scientific Advisory Board or as consultant to ATARA Biotherapeutics, Bayer, Carisma Therapeutics, Imugene, and Takeda Therapeutics, and has patents, royalties, and intellectual property on mesothelin-targeted CARs and other T cell therapies, and method for detection of cancer cells using virus, and pending patent applications on T cell therapies. D.E.D. is a cofounder and Chief Medical Officer of Theromics, Inc., a consultant to Boston Scientific and Perseon Medical, receives royalties from Springer Verlag and UpToDate.

Funding Information

The authors acknowledge the support of NIH Cancer Center Support Grant (P30 CA008748) for core laboratory services that were used for the presented work. G.S. acknowledges grant and funding support from the National Cancer Institute of the National Institutes of Health under Award Number R01CA236615, the Dept. of Defense CDMRP PRCRP Award CA170630 and CA190888, and the Institute for Applied Life Sciences in the University of Massachusetts at Amherst. P.S.A.'s laboratory work is supported by grants from the National Institutes of Health (P30 CA008748, R01 CA236615-01, and R01 CA235667), the U.S. Department of Defense (BC132124, LC160212, CA170630, and CA180889), the Batishwa Fellowship, the Comedy vs Cancer Award, the Dalle Pezze Foundation, the Derfner Foundation, the Esophageal Cancer Education Fund, the Geoffrey Beene Foundation, the Memorial Sloan Kettering Technology Development Fund, the Miner Fund for Mesothelioma Research, the Mr. William H. Goodwin and Alice Goodwin, the Commonwealth Foundation for Cancer Research, and the Experimental Therapeutics Center of Memorial Sloan Kettering Cancer Center.

Supplementary Material

Supplementary Figure S1