Systemic treatment options for patients with locally advanced or metastatic basal cell carcinoma (BCC) are limited, particularly when tumors are refractory to anti-programmed cell death protein-1 (PD-1). A better understanding of immune checkpoint expression within the BCC tumor microenvironment may inform combinatorial treatment strategies to optimize response rates. CD3, PD-1, programmed death ligand-1 (PD-L1), lymphocyte activation gene 3 (LAG-3), and T-cell immunoglobulin domain and mucin domain 3 (TIM-3)+ cell densities within the tumor microenvironment of 34 archival, histologically aggressive BCCs were assessed. Tumor infiltrating lymphocyte (TIL) expression of PD-1, PD-L1, and LAG-3, and to a lesser degree TIM-3, correlated with increasing CD3+ T-cell densities (Pearson’s r=0.89, 0.72, 0.87, and 0.63, respectively). 100% of BCCs (34/34) demonstrated LAG-3 and PD-1 expression in >1% TIL; and the correlation between PD-1 and LAG-3 densities was high (Pearson’s r=0.89). LAG-3 was expressed at ~50% of the level of PD-1. Additionally, we present a patient with locally-advanced BCC who experienced stable disease during and after 45 weeks of first-line anti-PD-1 (nivolumab), followed by a partial response after the addition of anti-LAG-3 (relatlimab). Longitudinal biopsies throughout the treatment course showed a graduated increase in LAG-3 expression after anti-PD-1 therapy, lending support for coordinated immunosuppression and suggesting LAG-3 as a co-target for combination therapy to augment the clinical impact of anti-PD-(L)1.

Supplemental Table 1. Clinical and pathologic characteristics of patients included in tissue microarrays Supplementary Figure 1. B7-H4 expression is low in human and mouse lung cancer cell lines in vitro Supplementary Figure 2. B7-H1 expression does not change with MAPK or proliferation inhibition Supplementary Figure 3. Scoring of PD-L1 and pS6 staining in lung adenocarcinoma and squamous cell carcinoma TMAs Supplementary Figure 4. Co-expression of pS6 and PD-L1 staining in TMAs Supplemental Figure 5. The combination of rapamycin and anti-PD-1 antibody decreases KRAS-driven lung tumor growth Supplementary Figure 6. The combination of rapamycin and αPD-1 blockade decreases NNK-derived syngeneic lung tumor growth.

Antibodies used for flow cytometry (clones and dilutions)

Cells were first gated on size and singularity by forward scatter and side scatter. Nonviable cells were excluded by live/dead gating. Live cells were gated on CD11c and CD11b, then on CD45 and F4/80 to identify macrophage and dendritic cell phenotypes. Finally myeloid cells were gated for expression of TIM-3. Final populations were (A) F4/80+CD45hiCD11c-CD11bhi (B) F4/80+CD45dimCD11bhi (C) F4/80+CD45hiCD11c+CD11b+ and (D) F4/80-CD45dimCD11c+CD11b-.

Melanocytic archival specimens analyzed for PD-L1 expression, BRAF mutational status, and inflammatory infiltrate. PD-L1 and MHC I induction by IFN-g on cultured human melanoma cell lines, according to BRAF codon 600 genotype.

Supplementary Table S4: DAVID clustering of 1660 Illumina probes up-regulated at least 2-fold in PD-L1+ melanomas

Supplementary Table and Figure Legends

PDF file - 685KB, Supplementary Figure 1 showing a photomicrograph of intratumoral lymphoid aggregates. Supplementary Figure 2 showing the association of PD-L1 expression with immune infiltrates. Supplementary Figure 3 showing the PD-L1 and PD-1 expression by the immune infiltrates in colorectal carcinoma. Supplementary Figure 4 showing the immunoarchitecture of a melanoma lymph node metastasis. Supplementary Table 1 showing the relationship between pre-treatment microenviornmental parameters in the tumor specimen demonstrating the highest expression of each variable across all specimens from a single patient and the clinical response to anti-PD-1.

Each mouse was individually imaged with cone beam computed tomography (CBCT) using the SARRP with a 65 kVp and 0.7 mA beam. Using the treatment planning system from the SARRP system (Muriplan) and the CBCT, the target was exactly placed 3mm below the skull's burr hole. The planning system calculated the x-ray beam's (220kVp and 13mA) time of exposure according to the prescribed dose and moved the motorized couch to its target location, after which a 3-mm beam centered on the burr hole and underlying tumor was used to administer a total of 10 Gy radiation per animal at a rate of 1.9 Gy/min (18). The isodose distribution is shown on the figure provided. The dose to adjacent organs or rest of the brain is insignificant as the beam's penumbra demonstrates complete drop off on the edge of the field (17, 50, 51)

Multiplex immunohistochemistry/immunofluorescence (mIHC/mIF) is a developing technology that facilitates the evaluation of multiple, simultaneous protein expressions at single-cell resolution while preserving tissue architecture. These approaches have shown great potential for biomarker discovery, yet many challenges remain. Importantly, streamlined cross-registration of multiplex immunofluorescence images with additional imaging modalities and immunohistochemistry (IHC) can help increase the plex and/or improve the quality of the data generated by potentiating downstream processes such as cell segmentation. To address this problem, a fully automated process was designed to perform a hierarchical, parallelizable, and deformable registration of multiplexed digital whole-slide images (WSIs). We generalized the calculation of mutual information as a registration criterion to an arbitrary number of dimensions, making it well suited for multiplexed imaging. We also used the self-information of a given IF channel as a criterion to select the optimal channels to use for registration. Additionally, as precise labeling of cellular membranes in situ is essential for robust cell segmentation, a pan-membrane immunohistochemical staining method was developed for incorporation into mIF panels or for use as an IHC followed by cross-registration. In this study, we demonstrate this process by registering whole-slide 6-plex/7-color mIF images with whole-slide brightfield mIHC images, including a CD3 and a pan-membrane stain. Our algorithm, WSI, mutual information registration (WSIMIR), performed highly accurate registration allowing the retrospective generation of an 8-plex/9-color, WSI, and outperformed 2 alternative automated methods for cross-registration by Jaccard index and Dice similarity coefficient (WSIMIR vs automated WARPY, P < .01 and P < .01, respectively, vs HALO + transformix, P = .083 and P = .049, respectively). Furthermore, the addition of a pan-membrane IHC stain cross-registered to an mIF panel facilitated improved automated cell segmentation across mIF WSIs, as measured by significantly increased correct detections, Jaccard index (0.78 vs 0.65), and Dice similarity coefficient (0.88 vs 0.79).

Gating strategy to assess for surface expression of TIM-3 on (A) CD4+ and (B) CD8+ T cells isolated from peripheral lymph nodes, lungs, livers, spleens, and brains.

Abstract Multiplex immunofluorescence (mIF) can detail spatial relationships and complex cell phenotypes in the tumor microenvironment (TME). However, the analysis and visualization of mIF data can be complex and time-consuming. Here, we used tumor specimens from 93 patients with metastatic melanoma to develop and validate a mIF data analysis pipeline using established flow cytometry workflows (image cytometry). Unlike flow cytometry, spatial information from the TME was conserved at single-cell resolution. A spatial uniform manifold approximation and projection (UMAP) was constructed using the image cytometry output. Spatial UMAP subtraction analysis (survivors vs. nonsurvivors at 5 years) was used to identify topographic and coexpression signatures with positive or negative prognostic impact. Cell densities and proportions identified by image cytometry showed strong correlations when compared with those obtained using gold-standard, digital pathology software (R2 > 0.8). The associated spatial UMAP highlighted “immune neighborhoods” and associated topographic immunoactive protein expression patterns. We found that PD-L1 and PD-1 expression intensity was spatially encoded—the highest PD-L1 expression intensity was observed on CD163+ cells in neighborhoods with high CD8+ cell density, and the highest PD-1 expression intensity was observed on CD8+ cells in neighborhoods with dense arrangements of tumor cells. Spatial UMAP subtraction analysis revealed numerous spatial clusters associated with clinical outcome. The variables represented in the key clusters from the unsupervised UMAP analysis were validated using established, supervised approaches. In conclusion, image cytometry and the spatial UMAPs presented herein are powerful tools for the visualization and interpretation of single-cell, spatially resolved mIF data and associated topographic biomarker development.

Next-generation tissue-based biomarkers for immunotherapy will likely include the simultaneous analysis of multiple cell types and their spatial interactions, as well as distinct expression patterns of immunoregulatory molecules. Here, we introduce a comprehensive platform for multispectral imaging and mapping of multiple parameters in tumor tissue sections with high-fidelity single-cell resolution. Image analysis and data handling components were drawn from the field of astronomy. Using this "AstroPath" whole-slide platform and only six markers, we identified key features in pretreatment melanoma specimens that predicted response to anti-programmed cell death-1 (PD-1)-based therapy, including CD163+PD-L1- myeloid cells and CD8+FoxP3+PD-1low/mid T cells. These features were combined to stratify long-term survival after anti-PD-1 blockade. This signature was validated in an independent cohort of patients with melanoma from a different institution.

Supplemental digital content is available in the text. Objectives The aim of this study was to characterize the tumor microenvironment of patients with gastroenteropancreatic neuroendocrine tumors relative to progression-free survival (PFS). Methods Immune profiling for CD3, CD8, programmed death-1/programmed death-ligand 1, and indoleamine 2,3-dioxygenase expression in 2 cohorts of gastroenteropancreatic neuroendocrine tumors: patients with short PFS (<4 years, n = 12) versus long PFS (≥4 years, n = 14) after surgery. Immune infiltrates in the tumor and interface were quantified. Programmed death-ligand 1 expression was determined within the tumor, stroma, and interface. Results Patients with shorter PFS had larger tumors (P = 0.02), mostly in the pancreas (P = 0.04). We observed a higher mean expression of CD3+, CD8+, programmed death-1+ cells, and indoleamine 2,3-dioxygenase at the interface compared with the tumor: log 10 mean differences 0.56 (95% confidence interval [CI], 0.43–0.68; P < 0.0001), 0.45 (95% CI, 0.32–0.59; P = 0.0002), 0.50 (95% CI, 0.40–0.61; P < 0.0001), and 0.24 (95% CI, 0.03–0.46; P = 0.046), respectively. Patients with longer PFS had higher intratumoral CD3+ T cells, log 10 mean difference 0.38 (95% CI, 0.19–0.57; P = 0.004). Programmed death-ligand 1 expression tended to be higher among patients with shortened PFS (odds ratio, 2.00; 95% CI, 0.68–5.91). Conclusions Higher intratumoral CD3+ T-cell infiltrate was associated with longer PFS after resection.

Multiplex IF (mIF) provides a detailed characterization of spatial relationships and complex cell phenotypes in the tumor microenvironment. However, the data-analysis and visualization is complex and time-consuming. Here, we developed a platform to analyze mIF data through flow cytometry workflows (image cytometry), while maintaining spatial information, and applied it to tissue microarrays of metastatic melanoma specimens (n=93; 6-plex mIF panel: PD-1, PD-L1, CD163, CD8, FoxP3, Sox10/S100). Then, we used a UMAP-based approach driven by cell-to-cell distances (rather than fluorescence intensity) to display and analyze geographic organization and cell interactions. Our pipeline provided equivalent results to the digital pathology gold standard with faster run times (5-fold reduction) and higher reproducibility. We identified key prognostic immune variables, including CD8PD1Low and CD8PD1Neg cells which associated with longer overall survival (OS, both p<0.01), and CD163PDL1Neg cells which associated with shorter OS (p=0.001). The spatial UMAPs showed that PD-L1 and PD-1 intensities were spatially encoded, and their expression on distinct cell subsets was organized in geographic clusters. Specifically, PD-L1Hi cells co-located to areas of CD8 cells, and PD-1Hi cells were observed near dense collections of tumor cells. Spatial UMAP subtraction analysis (survivors vs. non-survivors at 5 years) identified geographic and co-expression signatures associated with improved prognosis, i.e. CD8-driven PD-L1 expression and lacking CD163PDL1Neg macrophages. These data demonstrate the use of image cytometry and spatial UMAPs for improved visualization and interpretation of single-cell, spatially-resolved mIF data.

Abstract Background: Multidimensional, spatially resolved analyses of immune and tumor cells within the TME of patients treated with checkpoint inhibitors will provide clinically translatable mechanistic insights and potentiate biomarker discovery. To achieve this goal, information from pathology specimens needs to be captured at a single cell level with high fidelity and in meaningfully sized cohorts. To date, efforts have been limited by inadequate tissue sampling and previously unrecognized errors in staining, imaging and data analysis. Here we describe the ‘AstroPath' platform, where strategies from the field of astronomy were adapted to study pathology specimens and generate large high quality mIF data. Methods: Potential error was identified and addressed at each stage of 6-plex (PD-1, PD-L1, FoxP3, CD163, CD8, tumor marker) mIF assay development. Whole slides from formalin-fixed paraffin embedded tissue specimens were stained with the optimized assay and imaged using a multispectral microscope (Vectra 3.0) with 20% overlap of high power fields (HPFs). The overlaps were used to quantify and correct optical lens distortion, HPF alignment, and illumination variation. Errors from cell segmentation algorithms, batch-to-batch staining variation, and HPF sampling were also addressed. The resultant mIF data were organized and analyzed using a large, relational database. Results: The optimized mIF assay captured equivalent signal compared to gold standard chromogenic immunohistochemistry and 2x more signal for PD-1, PD-L1 and FoxP3 compared to the manufacturer's recommended protocol. Errors and corrections for imaging included: pixel alignment error reduced from ~10 to <0.5 pixels at edges of HPFs; illumination variation reduced from 10% to 3% per HPF; over-counting of larger cells, e.g. tumor cells, reduced by ~25% using custom cell segmentation and ‘multi-pass' phenotyping algorithms; and batch-to-batch variation reduced by ~50% by normalizing to tissue controls. Correction of these errors that would otherwise be compounded at each stage, allowed for more accurate and reliable cell type and marker intensity comparisons across samples. Lastly, the entire slide rather than select HPFs were imaged, resulting in ~100x more HPFs analyzed per slide. The whole slide imaging approach also corrected for other potential source of errors, i.e., sampling error due to tumor heterogeneity and operator-dependent field selection. Conclusion: Here we present an end-to-end pathology workflow with rigorous quality control for creating quantitative, spatially resolved mIF datasets using lessons derived from the field of astronomy. Such approaches will vastly improve standardization and scalability of mIF technologies, enabling cross-site comparisons and eventual clinical translation as biomarker discovery platforms or standard diagnostic tests. Citation Format: Sneha Berry, Nicolas Giraldo, Benjamin Green, Elizabeth Engle, Haiying Xu, Aleksandra Ogurtsova, Daphne Wang, Julie E. Stein, Peter Nguyen, Suzanne Topalian, Angelo DeMarzo, Drew M. Pardoll, Robert A. Anders, Tricia R. Cottrell, Alexander S. Szalay, Janis M. Taube. The ‘AstroPath' platform for spatially resolved, single cell analysis of the tumor microenvironment (TME) using multispectral immunofluorescence (mIF) [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6584.

After publication of this supplement [1, 2], it was brought to our attention that due to an error authors were missing in the following abstracts. This has now been included in this correction.

BACKGROUND:Programmed cell death ligand 1 (PD-L1) is a transmembrane glycoprotein that interacts with the receptor programmed cell death 1 (PD-1) to suppress T-cell activation, reduce adjacent tissue damage, and promote tolerance to self-antigens. Tumors may express PD-L1 as a mechanism to evade immune detection. Recent clinical trials have demonstrated the efficacy of PD-L1/PD-1 antagonists through activation of tumor-infiltrated CD8+ T cells. The aim of this study was to determine the expression pattern of PD-L1 and PD-1 in olfactory neuroblastoma (ONB) tumor cells and to determine the presence of PD-1+ and CD8+ lymphocytes in the ONB immune microenvironment. METHODS:Immunohistochemistry for expression of PD-L1, PD-1, and CD8 was performed on paraffin-embedded ONB tissue. RESULTS:Of the 10 primary site ONB samples, 4 demonstrated positive PD-L1 expression. Of PD-L1+ tumors, the 2 highest expressing samples were found to contain PD-1+ tumor cells. Of the 4 available metastatic samples, all of which arose from PD-L1- primary site ONB, 3 were positive for PD-L1 and contained PD-1+ tumor cells. PD-L1+ primary and metastatic tumors also demonstrated increased PD-1+ infiltrating lymphocytes in the tumor and stroma (11.6- and 4.62-fold increase) compared with PD-L1- samples (P < 0.05 and P = 0.068 respectively). PD-L1+ specimens demonstrated increased CD8+ lymphocytes in the tumor and stroma (7.46- and 2.14-fold increase) compared with PD-L1- tumors (P < 0.05 for both). CONCLUSIONS:These data demonstrate that a proportion of ONB primary and metastatic tumors express PD-L1 and possess an associated tumor and stromal infiltrate of PD-1+ and CD8+ lymphocytes.

Inflammatory myofibroblastic tumor is a rare mesenchymal tumor occurring at many anatomic sites, with a predilection for children and young adults. Often indolent, they can be locally aggressive and can metastasize, resulting in significant morbidity and mortality. Therapeutic options are often limited. The identification of underlying kinase mutations has allowed the use of targeted therapy in a subset of patients. Unfortunately, not all tumors harbor mutations and resistance to tyrosine kinase inhibitor therapy is a potential problem. We hypothesized that these tumors may be amenable to PD-L1 therapy given the immune nature of the tumor. PD-L1 expression in inflammatory myofibroblastic tumors has not yet been defined. The purpose of this study was to explore PD-L1 expression in inflammatory myofibroblastic tumors, as adaptive PD-L1 expression is known to enrich for response to anti-PD-1/PD-L1 therapies. Expression of PD-L1 (clone SP142) was assessed in 35 specimens from 28 patients. Positivity was defined as membranous expression in ≥5% of cells and evaluated separately in tumor and immune cells. Adaptive vs. constitutive patterns of tumor cell PD-L1 expression were assessed. PD-L1 status was correlated with clinicopathologic features. CD8+ T cell infiltrates were quantified by digital image analysis. ALK status was assessed by immunohistochemistry and/or FISH. Twenty-four (69%) tumors had PD-L1(+) tumor cells and 28 (80%) showed PD-L1(+) immune cells. Most recurrent and metastatic tumors (80%) and ALK(−) tumors (88%) were PD-L1(+). Adaptive PD-L1 expression was present in 23 (96%) of PD-L1(+) tumors, which also showed a three–four fold increase in CD8+ T cell infiltration relative to PD-L1(−) tumors. Constitutive PD-L1 expression was associated with larger tumor size (p = 0.002). Inflammatory myofibroblastic tumors show frequent constitutive and adaptive PD-L1 expression, the latter of which is thought to be predictive of response to anti-PD-1. These data support further investigation into PD-1/PD-L1 blockade in this tumor type.