Whole exome sequencing

The protein-coding sequences of a gene are called exons, and all the combined exons in a genome are referred to as the exome. With existing technology, 95% of the human exome can be sequenced. Therefore, the term ‘exome sequencing’ is more accurate than the term ‘whole-exome sequencing (WES)’. It is noteworthy that while the human exome comprises all coding nuclear DNA sequences, mitochondrial DNA is not included. The exome represents less than 2% of the human genome, but contains about 85% of known disease-related variants, establishing exome sequencing as a cost-effective alternative to whole genome sequencing (WGS). Exome sequencing using exome enrichment can efficiently identify coding variants across a wide range of applications, including population genetics, genetic disease, and cancer studies. However, exome sequencing techniques have non-standardized, highly variable coverage, including regions of the exome that are refractory to being accurately sequenced, such as genes containing a pseudogene, highly repetitive coding regions, large deletions, and duplications. Therefore, it is likely that some clinically significant mutations would be missed by exome sequencing due to inefficient capture of relevant exons:

• Targeted capture and massively parallel sequencing of 12 human exomes.189

• Educational materials—genetic testing: current approaches.190

Cancer is driven by genomic events, and different sets of genetic aberrations can characterize individual cancers. The use of high-throughput sequencing, including exome sequencing, to identify those changes can guide the identification of effective therapies currently available or still in clinical trials.

• Personalized oncology through integrative high-throughput sequencing: a pilot study.191

• Whole-exome sequencing of metastatic cancer and biomarkers of treatment response.192

Clinical implementation of genomic data to inform therapy necessitates that clinicians interpret the patient’s genomic profile, including both tumor and germline DNA. Currently, only a limited number of genomic markers in specific cancer settings have shown strong evidence of differential response to specific therapies, as these are targetable mutations with approved or investigational therapies. For many other known genomic alterations, there are no data or insufficient data to support routine clinical implementation of biomarker-based therapy.

• A decision support framework for genomically informed investigational cancer therapy.193

• Precision oncology in the age of integrative genomics.194

TCR sequencing and clonality

Human T cells mature in the thymus from hematopoietic progenitors, gain the ability to recognize foreign antigens, and provide protection against a vast array of pathogens.

A complex molecular mechanism in T cells based on somatic recombination leads to the expression of highly polymorphic surface receptors, the TCRs, and provides the immune system with functional plasticity. TCR sequencing (TCR-seq) produces large numbers of short DNA sequences covering key regions of the TCR coding sequence, allowing quantification of T cell diversity at high resolution. Reduced cost of high-throughput sequencing technologies has enabled the identification of immune response signatures based on sequence analysis. In consequence, high-throughput TCR-seq has been established as a tool to analyze antigen specificity, clonality, and diversity of T lymphocytes:

• Linking T-cell receptor sequence to functional phenotype at the single-cell level.195

• Overview of methodologies for T-cell receptor repertoire analysis.196

The characterization of the TCR repertoire through TCR-seq is of great scientific and potential clinical relevance because it accurately describes T cell dynamics in a wide range of diseases, including infection, autoimmune diseases, and malignancies:

• Tumor-infiltrating lymphocytes in colorectal tumors display a diversity of T cell receptor sequences that differ from the T cells in adjacent mucosal tissue.197

• TCR sequencing facilitates diagnosis and identifies mature T cells as the cell of origin in CTCL.198

• Characteristics of tumor infiltrating lymphocyte and circulating lymphocyte repertoires in pancreatic cancer by the sequencing of T cell receptors.199

• A new high-throughput sequencing method for determining diversity and similarity of T cell receptor (TCR) α and β repertoires and identifying potential new invariant TCR α chains.200

High-throughput TCR-seq is rapidly evolving, and numerous validated procedures for clonotype identification and TCR repertoire analysis exist. However, no gold standard method has been established in the field. A number of platforms are available, including DNA-based (eg, Adaptive), RNA-based (eg, iRepertoire), bulk TCR-seq, and single-cell TCR-seq (10X) technologies. Different approaches may be more applicable than others for different scientific purposes, but can be subject to possible method-specific biases. There are innovative approaches, such as combining TCR-seq with an assay for transposase-accessible chromatin analysis at the single cell level for information on TCR specificity and the epigenomic state of individual T cells, which will likely expand the academic and clinical utility of this technology:

• Single cell T cell receptor sequencing: techniques and future challenges.201

• Transcript-indexed ATAC-seq for precision immune profiling.202

• Quantifiable predictive features define epitope-specific T cell receptor repertoires.203

• Identifying specificity groups in the T cell receptor repertoire.204

• Using T cell receptor repertoires to understand the principles of adaptive immune recognition.205

Epigenetic immune cell quantification with qPCR-based assisted cell counting (qPACC)

DNA-based, immune cell subset-specific epigenetic markers have recently been identified and can be used to differentiate leukocytes, lymphocytes, and other cell types of interest. These markers can also be used for epigenetic cell counting.

• DNA demethylation in the human FOXP3 locus discriminates regulatory T cells from activated FOXP3⁺ conventional T cells.206

• Quantitative DNA methylation analysis of FOXP3 as a new method for counting regulatory T cells in peripheral blood and solid tissue.207

• Epigenetic quantification of tumor-infiltrating T-lymphocytes.208

Relative cell numbers can be quantified using quantitative real-time PCR (qPCR) based on knowledge of unmethylated DNA regions of previously characterized cell types, in combination with bisulfite conversion (BSC). During BSC, unmethylated cytosines in DNA convert to uracil, but methylated cytosines are protected and remain unchanged. The resulting sequence changes are the foundation for developing differentiating primer and probe sets for qPCR on clinical samples. Different cell types relevant for immune monitoring during immunotherapy have been described, for example, Treg, CD3⁺, CD4⁺, and CD8⁺ T cells, B cells, natural killer (NK) cells, and neutrophils. In addition, several approaches to control, calibration, and quantification are used that allow the calculation of immune cell concentrations with qPCR-based assisted cell counting (qPACC), in a process referred to as epigenetic cell counting.

• Quantitative real-time PCR assisted cell counting (qPACC) for epigenetic-based immune cell quantification in blood and tissue.15

• Epigenetic immune cell counting in human blood samples for immunodiagnostics.209

The general stability of DNA as well as its methylation, in addition to the small sample volume needed, provide epigenetic-based assays the advantage of being less susceptible to challenges related to sample amount and quality, and permit the measurement of different immune cell subset frequencies without the need to count intact cells.

The immune cell subset-specific epigenetic markers are developed based on highly purified, fluorescence-activated cell sorting (FACS)-sorted cells, obtained from whole blood of healthy donors. Cancer is a disease affecting the DNA of patients, including, but not limited to, DNA strand breaks, gene duplication, and aberrant DNA methylation. In the immuno-oncology setting, this represents a theoretical obstacle to the successful use of this technology, since the specificity of the established cell type markers from healthy individuals could be compromised in cancer, precluding their use, at least in tumor-affected tissues.

• Demethylation of the FOXP3 gene in human melanoma cells precludes the use of this epigenetic mark for quantification of Tregs in unseparated melanoma samples.210

Conclusions

Studies have demonstrated the influence of the microbiome on carcinogenesis and response to therapy. However, it will require extensive research and resources to obtain reliable and clinically actionable information. The microbiota can vary considerably over time, between individuals, and in different areas of the body. Establishing clear distinctions in regard to the cause and effect of tumor-associated microbiota and concurrent changes in the microbiota is essential. The prognostic potential and theoretical value of therapeutic intervention toward the microbiota offer the exciting possibility of new tools to fight cancer, while making current therapies more effective and reducing side effects.

Conclusions

The use of MitoChip to track mutations in mtDNA has been shown to be relevant in diverse cancer settings. However, its diagnostic, prognostic, and clinical value is still debated in the field based on several technology-related obstacles to standardization and relevant controls. New NGS approaches, combined with sophisticated data analysis to select mutations with likely functional relevance, could help to more systematically evaluate the potential role of mtDNA mutations in tumor biology in the future.

ATAC-seq

Eukaryotic DNA is extensively packaged around histone proteins, forming nucleosomes, which are condensed into higher levels of packaging to allow chromatin to fit within the nucleus of a cell. Nevertheless, processes that allow ‘open’ chromatin states, which permit transcription factors and histone post-translational changes to influence gene expression, orchestrate active transcription.

Assay for transposase accessible chromatin with high-throughput sequencing, or ATAC-seq for short, is a method for mapping chromatin accessibility genome-wide. Hyperactive transposase Tn5 is used to cut and ligate adaptors for high-throughput sequencing of DNA in regions of high accessibility. This allows mapping of ‘open’ chromatin areas as well as nucleosome topology. The following paper describes ATAC-seq of lymphoblastoid cells.

• ATAC-seq: a method for assaying chromatin accessibility genome-wide.235

Significant chromatin heterogeneity can exist within a population of cells. Accuracy is enhanced when chromatin assays can be performed at a single cell level. The following paper talks about single-cell chromatin profiling.

• A rapid and robust method for single cell chromatin accessibility profiling.236

The profiling of chromatin of different cells within a cell population can inform cell subsets. The following papers describe single-cell ATAC-seq (scATAC-seq) in multiple cells to cluster them.

• Single-cell ATAC-seq: strength in numbers.237

• High-throughput chromatin accessibility profiling at single-cell resolution.238

Transcript-indexed ATAC-seq is a tool by which the TCR gene is sequenced along with ATAC-seq at the single cell level. Transcript-indexed ATAC-seq enables an analysis of the epigenetic landscape of a clonal T cell population and also enables discovery of regulatory pathways affecting T cell function.

• Transcript-indexed ATAC-seq for precision immune profiling.202

In this paper, the team has identified a programmable and a dysfunctional chromatin state in tumor-infiltrating T cells based on chromatin assays. They have identified the epigenetic processes associated with dysfunctional immune cells and surface biomarkers to identify reprogrammable T cells.

• Chromatin states define tumour-specific T cell dysfunction and reprogramming.239

Studies of chromatin states using ATAC-seq in immunology to reveal epigenetic heterogeneity, the mechanistic basis of T cell dysfunction, and distinct T cell subsets are as follows.

• Joint single-cell DNA accessibility and protein epitope profiling reveals environmental regulation of epigenomic heterogeneity.240

• Newly identified T cell subsets in mechanistic studies of food immunotherapy.241

Satpathy AT et al 242 describe the use of scATAC-seq with a droplet-based method on a widely used single-cell 10X Chromium platform to discover cell types and regulatory DNA elements in complex tissues. They performed scATAC-seq using bone marrow and blood samples to characterize the chromatin landscape of cell subtypes and their differentiation trajectories. They then performed scATAC-seq on primary tumor tissue before and after treatment with PD-1 blockade. The authors demonstrate the ability to deconvolute the TME at the single cell level, revealing subpopulations of immune cells and malignant cells.

• Massively parallel single-cell chromatin landscapes of human immune cell development and intratumoral T cell exhaustion.242

The following paper describes the various methods used to assess chromatin accessibility, including the limitations, advantages, and specimen requirements of each technique: micrococcal nuclease sequencing (MNase-seq), DNase-seq, formaldehyde-assisted isolation of regulatory elements sequencing (FAIRE-seq), and ATAC-seq.

• Chromatin accessibility: a window into the genome.243

ChIP arrays (ChIP on chip) and ChIP-seq

The interaction of DNA with transcription factors and histones affects gene expression and cell phenotype. Chromatin immunoprecipitation (ChIP) is used widely to establish specific DNA–protein interactions. When ChIP is combined with whole genome DNA microarrays, the assay is known as a ChIP microarray or ChIP on chip. A genome-wide assessment of protein–DNA interactions was made possible through ChIP arrays and led to the discovery of epigenomic transcriptional regulation. Several tiling microarray platforms for common model organisms were commercially developed, and bioinformatics tools were generated to analyze data from those platforms. Some early descriptions of methods, applications, and analytical tools for ChIP on chip are referenced in the following.

• ChIP-chip: considerations for the design, analysis, and application of genome-wide chromatin immunoprecipitation experiments.244

• Chromatin immunoprecipitation for determining the association of proteins with specific genomic sequences in vivo.245

• Genome-wide profiling of PPARγ:RXR and RNA polymerase II occupancy reveals temporal activation of distinct metabolic pathways and changes in RXR dimer composition during adipogenesis.246

• rMAT-- an R/Bioconductor package for analyzing ChIP-chip experiments.247

More recently, massive parallel sequencing of DNA fragments crosslinked to protein has been made possible by a newly developed high-throughput sequencing technology. This technology, referred to as ChIP-seq, represents a large advancement in the study of DNA–protein interactions.

ChIP-seq has higher sensitivity and specificity than ChIP on chip and can be used to analyze any sequenced species, since it is not dependent on a microarray. It is also more cost-effective than ChIP on chip. The use of ChIP-seq in non-coding regions of the DNA has the potential to identify the biological role of single nucleotide polymorphisms (SNPs) associated with a disease state.

The following study compares ChIP on chip using the Agilent tiling microarray with ChIP-seq using the Illumina GAII.

• ChIP-chip vs ChIP-seq: lessons for experimental design and data analysis.248

The following describe ChIP-seq assays and bioinformatic tools for data analysis:

• Practical guidelines for the comprehensive analysis of ChIP-seq data.249

• ChIP-seq and beyond: new and improved methodologies to detect and characterize protein-DNA interactions.250

• Identifying and mitigating bias in next-generation sequencing methods for chromatin biology.251

Methylation arrays

In the human body, using the same genome, at least 200 distinct cell phenotypes can be created by means of epigenomic processes. An epigenetic imprint of cell types enhances cell recognition. In fact, the recent years have seen the identification of cell types based on epigenomic profiles by making use of reference methylation profiles of known cell types from the gene expression omnibus (GEO). In one such study, investigators performed epigenetic deconvolution of SCCHN samples from the TCGA, identifying different cell types in the tumors by histoepigenetic means. By identifying various cell populations in the tumors as immune, cancer, and epidermal cells, the investigators were able to epigenetically profile the distinct types of SCCHN by differential methylation analysis:

• Histoepigenetic analysis of HPV- and tobacco-associated head and neck cancer identifies both subtype-specific and common therapeutic targets despite divergent microenvironments.252

Several methylated genes serve as biomarkers of disease diagnosis, prognosis, and therapeutic targeting. With the availability of epigenome-wide methylation arrays, a much more amplified biomarker discovery platform is available.

The computational analysis of DNA methylation data involves multiple steps of data processing. The following papers describe the use of Illumina-based Infinium DNA methylation BeadChip assays, which are considered to be the gold standard method for DNA methylation analysis. These assays provide quantitative measurement of DNA methylation levels in a predetermined set of cytosine residues using a microarray format. With advancing capabilities and demands, and increasing numbers of target CpG sites, the kits have advanced from HumanMethylation 27K BeadChip (27K array) to HumanMethylation 450K (450K array) and more recently the Infinium MethylationEPIC (850K array). The following reviews describe the steps involved in data acquisition and processing.

• Review of processing and analysis methods for DNA methylation array data.253

• Computational and statistical analysis of array-based DNA methylation data.254

In a multicenter study based in Europe, investigators have successfully identified and validated a DNA methylation signature, termed EPIMMUNE, associated with clinical benefit from anti-PD-1 blockade in patients with stage IV lung cancer. Among the methylated genes, forkhead box P1 (FOXP1) was confirmed to be a predictor of clinical benefit from anti-PD-1 therapies. The epigenetic signature was not associated with the PD-L1 status, mutational load, or CD8 immunostaining.

• Epigenetic prediction of response to anti-PD-1 treatment in non-small-cell lung cancer: a multicentre, retrospective analysis.255

When compared with studies that require live cells for immune subset quantification and monitoring, assessing gene methylation by epigenetic assays to identify cell subsets is a more feasible approach. The same principle can be used to characterize the cell population of tumor tissue as well.

• Novel technologies and emerging biomarkers for personalized cancer immunotherapy.7

PBMCs and T cells from patients with HCC have been shown to have a DNA methylation signature distinct from those who do not have HCC. This methylation signature in the host immune cells deepens as the HCC advances.

• The signature of liver cancer in immune cells DNA methylation.256

microRNA arrays

Overwhelming portions of the human DNA do not encode protein. The RNA resulting from the transcription of approximately 98% of DNA is non-coding RNA (ncRNA). In the early 2000s, the role of ncRNAs as they affect translation of proteins started to emerge. Further study revealed a large network of interactions between ncRNA and the cell physiological machinery. This has led to a new understanding of what used to be considered ‘junk’ DNA and RNA. These are now recognized as important players in the regulation of the cell machinery and as oncogenic drivers and suppressors.

• Non-coding RNA networks in cancer.257

microRNA (miRNA) are 22-nucleotide short RNA molecules that are highly conserved through evolution. miRNA binds to mRNA transcripts of a gene, which leads to cleavage of the mRNA or shutdown of gene translation. miRNAs are active participants in normal cellular, developmental, and host environment processes such as intercellular communication through exosomes. Additionally, these tiny molecules play a fascinating role in cancer, as they have the powerful ability to orchestrate carcinogenesis and therapy resistance in cancer. Some miRNAs may stimulate an oncogenic process directly by acting as a ligand, termed onco-miRNAs (such as miR-21/miR-29a), while others such as miR34a act as tumor suppressors. For a detailed review on this subject, refer to the following.

• The role of microRNAs in human cancer.258

Circulating and exosomal miRNAs can be used as diagnostic biomarkers for diseases such as cancer. The following reviews discuss the process of developing miRNAs as biomarkers, various profiling platforms, sample preparation, and analytical strategies.

• MicroRNA as biomarkers and diagnostics.259

• Potential pitfalls in microRNA profiling.260

The following reviews outline the role of miRNAs in mediating immune responses by influencing cellular signaling in immune cells, influencing both innate and adaptive immunity.

• MiRNAs: dynamic regulators of immune cell functions in inflammation and cancer.261

• Nuclear functions of mammalian microRNAs in gene regulation, immunity and cancer.262

• Extracellular RNAs: a secret arm of immune system regulation.263

• miRNA regulation of innate immunity.264

• MicroRNAs as regulatory elements in immune system logic.265

The following is an excellent review of the role of miRNAs in oncogenesis and therapy resistance in melanoma. The authors describe an miRNA signature that affects angiogenic and inflammatory pathways and predicts resistance to both targeted and immunotherapy, highlighting this as a biomarker of resistance and a potential therapeutic target.

• MicroRNA-driven deregulation of cytokine expression helps development of drug resistance in metastatic melanoma.266

Studies that successfully sought and identified miRNA biomarker signatures in relation to immunotherapy in a variety of settings are listed below.

• Identification of a microRNA signature in dendritic cell vaccines for cancer immunotherapy.267

• Circulating immune cell and microRNA in patients with uveal melanoma developing metastatic disease.268

miRNA can also be used as therapeutic targets.

• Sequence-specific knockdown of EWS-FLI1 by targeted, nonviral delivery of small interfering RNA inhibits tumor growth in a murine model of metastatic Ewing’s sarcoma.269

• Evidence of RNAi in humans from systemically administered siRNA via targeted nanoparticles.270

Conclusions

Epigenetic biomarker discovery has an expansive reach in clinical practice for diagnostic, prognostic, and therapeutic purposes. Multiple epigenetic processes such as DNA methylation, histone post-translational changes, chromatin remodeling, and ncRNA production have been described in the context of normal tissue development, cancer, and immunology. Sample processing and preservation, miniaturization, and automation have improved the ability to perform large-scale reproducible epigenome-wide biomarker assays to define the epigenetic phenotype of cells. In conjunction with refined bioinformatics analytical processes, epigenetics has been successfully applied toward biomarker discovery in immunology. A shared knowledge of a growing epigenomic database, developed in conjunction with a much larger and more mature genomic database, facilitates establishment of epigenomic signatures of cell subsets, drug resistance, and other immunological biomarkers. Epigenetic changes can serve as biomarkers of diagnosis and prognosis, and are attractive therapeutic targets.

RNA sequencing

With the advent of NGS, RNA-seq has become mainstream in transcriptome analysis spanning basic and translational research. This methodology enabled sequencing and quantification of the transcriptional portraits of individual cells or thousands of samples, linking cellular and molecular phenotypes. Below is the first article that reported RNA-seq.

• Mapping and quantifying mammalian transcriptomes by RNA-seq.271

The following articles provide a high-level overview of RNA-seq, summarizing advantages over existing transcriptomic platforms and challenges. Technological advancements are covered, including improvements in transcription start site mapping, strand-specific measurements, gene fusion detection, small RNA characterization, detection of alternative splicing events, direct RNA sequencing, and approaches that enable profiling of small RNA quantities. Also reviewed are methods and tools developed for preprocessing high-throughput RNA-seq data and the analysis of differential gene expression.

• RNA-seq: a revolutionary tool for transcriptomics.272

• RNA sequencing: advances, challenges and opportunities.273

• From RNA-seq reads to differential expression results.274

These articles outline a historical timeline of transcriptomics, summarize various protocols and computational tools for RNA-seq, discuss the potential clinical utility of transcriptomic approaches, provide a useful toolbox with resources to analyze cancer transcriptomics, outline the lack of appropriate reference standards for validating RNA-seq, and illustrate the overabundance of competing computational tools.

• Cancer transcriptome profiling at the juncture of clinical translation.275

• Translating RNA sequencing into clinical diagnostics: opportunities and challenges.276

• Reference standards for next-generation sequencing.277

This article provides links to resources (Ensembl Compara, Gencode, Mouse Genomes Project, Mouse Phenome Database, OMIM (Online Mendelian Inheritance in Man), Rfam (RNA families database), Blueprint, ENCODE (Encyclopedia of DNA Elements), FANTOM (Functional Annotation of the Mammalian Genome), GTEx (Genotype-Tissue Expression) project, Human Cell Atlas Consortium, and so on) to help interrogate human and mouse transcriptomics.

• Comparative transcriptomics in human and mouse.278

EdgeSeq

EdgeSeq is a gene expression analysis platform developed by HTG which combines quantitative nuclease protection assay technology with NGS, using small amounts of starting material and delivering reproducible GEPs from poor-quality formalin-fixed paraffin-embedded (FFPE) tissue samples including haematoxylin and eosin (H&E)-stained tumor specimens. The EdgeSeq platform is capable of generating reliable expression data for thousands of genes from as little as 1 mm² FFPE tissue and crude FFPE tissue lysates equivalent to surface areas as low as 0.31 mm² of a 5 mm section, and can be particularly useful to interrogate biomarkers in oncology clinical trials, which often lack a sufficient amount of high-quality tumor tissue for other techniques.

• Reliable gene expression profiling from small and hematoxylin and eosin-stained clinical formalin-fixed, paraffin-embedded specimens using the HTG EdgeSeq platform.279

• EMT- and stroma-related gene expression and resistance to PD-1 blockade in urothelial cancer.168

Conclusions

RNA-seq is an open platform technology that has a number of potential advantages over gene expression microarrays, including an increased dynamic range of expression, measurement of focal changes (such as single nucleotide variants, insertions, and deletions), detection of different transcript isoforms, splice variants, and chimeric gene fusions (including previously unidentified transcripts and/or RNA species such as circular RNAs), and application to samples obtained from any biological species. As the cost of RNA-seq continues to decrease, this platform will most likely replace many applications focused on the analysis of transcriptome structure and dynamics. RNA-seq-based assays also have the potential to become a diagnostic platform in different therapeutic areas, including oncology. However, the establishment of appropriate quality standards and the adoption of best practices will be necessary to transform this exciting technology from a purely exploratory tool into a reliable diagnostic platform.

Conclusions

Single-cell transcriptomic analysis is rapidly transforming the field of biomedical research. While the promise and potential of this technology are apparent, several challenges remain. Efforts are underway to improve single-cell partitioning and whole transcriptome amplification, and to increase the sensitivity of scRNA-seq, which would allow detection of low-abundance RNAs and rare cells in the presence of biological and technical noise. An appropriate sample size and measurement of sufficient numbers of single-cell events are equally important to increase the accuracy and precision of scRNA-seq analyses. There is also a need for standardized scRNA-seq protocols, harmonized computational pipelines (including methods capable of resolving spatial single-cell transcriptomics), integrated single-cell data across experiments or modalities, and repositories dedicated to the massive and constantly increasing amounts of scRNA-seq data.

Transcriptome profiling with microarrays

While commonly used DNA arrays have become quickly outdated with the advent of RNA-seq, more advanced gene expression microarrays (eg, Affymetrix Human Transcriptome Array V.2.0 (HTA V.2.0), Clariom D and S arrays), which leverage the latest transcriptome data from multiple databases, are simple and fast tools for whole-transcriptome expression profiling and biomarker discovery. The Clariom D array is based on differential exon usage resulting from alternative splicing; this platform may be advantageous for the assessment of millions of distinct sequences and could be particularly useful in detecting and quantifying low abundance transcripts, or rare alternative splice variants. It is also worth noting that Clariom D arrays require very low RNA input and are compatible with formalin-fixed biological samples.

• https://www.thermofisher.com/us/en/home/life-science/microarray-analysis/transcriptome-profiling-microarrays/arrays-rna-seq.html 305

HTA V.2.0 was used for gene expression profiling to allow analysis of coding as well as non-coding and alternatively spliced transcripts in peripheral T cells isolated from patients with melanoma treated with anti-PD-1, anti-CTLA-4, and combinations of both antibodies. The results of this analysis revealed quantitatively and qualitatively distinct gene expression signatures associated with monotherapy or combination treatment.

• Combination therapy with anti-CTLA-4 and anti-PD-1 leads to distinct immunologic changes in vivo.176

nCounter and Digital Spatial Profiler

The nCounter gene expression system developed by NanoString enables enumeration of individual mRNAs using unique barcoding technology. Advantages over existing platforms include direct measurement of mRNA expression levels without enzymatic reactions, sensitivity coupled with high multiplex capability (up to 800 transcripts), and digital readout. Because detection probes in the nCounter analysis target relatively short mRNA sequences, this platform demonstrates outstanding performance in FFPE tissue samples and can be run on both purified RNA and tissue/cell lysates. It has also demonstrated utility in preclinical research using different species (mouse, rhesus/cynomolgus monkey) and sample types (tumor tissue, whole blood).

• Direct multiplexed measurement of gene expression with color-coded probe pairs.306

• The CDK4/6 inhibitor abemaciclib induces a T cell inflamed tumor microenvironment and enhances the efficacy of PD-L1 checkpoint blockade.307

• A conserved transcriptional response to intranasal Ebola virus exposure in nonhuman primates prior to onset of fever.308

• Evaluating robustness and sensitivity of the NanoString technologies nCounter platform to enable multiplexed gene expression analysis of clinical samples.309

The nCounter platform has been used to interrogate predictive and pharmacodynamic biomarkers in immuno-oncology clinical trials.

• IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade.164

• Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy.118

• Analysis of immune signatures in longitudinal tumor samples yields insight into biomarkers of response and mechanisms of resistance to immune checkpoint blockade.143

Digital Spatial Profiler (DSP) is another platform developed by NanoString aimed at assisting with positional information about mRNA/protein expression in a tissue sample. DSP is essentially a variation of multiplex mRNA in situ hybridization and IHC, but is based on the nCounter barcoding technology. The platform is capable of providing spatially resolved, digital characterization of proteins or mRNA in a highly multiplexed (up to 1000-plex) assay. The following two articles exemplify the translational utility of the DSP assay.

• Neoadjuvant immune checkpoint blockade in high-risk resectable melanoma.310

• Neoadjuvant vs adjuvant ipilimumab plus nivolumab in macroscopic stage III melanoma.311

QuantiGene Plex

The QuantiGene Plex (QGP) gene expression assay combines branched DNA (bDNA) technology with the Luminex fluorescent microbead-based platform. It uses cooperative hybridization, which allows for an exceptionally high degree of assay specificity by using multiple probes that hybridize to the same gene. The QGP assay is amenable to high-throughput analysis in a 96-well or 384-well format and is capable of multiplexing up to 80 targets in one well. Similar to nCounter, this platform is compatible with formalin-fixed material or samples with degraded RNA and can be run on both purified RNA and tissue/cell lysates.

• A multiplex branched DNA assay for parallel quantitative gene expression profiling.312

• QuantiGene Plex represents a promising diagnostic tool for cell-of-origin subtyping of diffuse large B-cell lymphoma.313

• The CDK4/6 inhibitor abemaciclib induces a T cell inflamed tumor microenvironment and enhances the efficacy of PD-L1 checkpoint blockade.307

High-throughput quantitative PCR

The microfluidics-based Biomark HD (Fluidigm) is a moderate/high-throughput qPCR system which is capable of analyzing the expression of multiple genes across multiple samples (up to 96×96) on a single plate with an integrated fluidic circuit format. Compared with plate-based high-throughput qPCR platforms, the Biomark HD provides excellent flexibility coupled with time-effectiveness and cost-effectiveness to explore a range of transcriptomic biomarkers in clinical samples. It is capable of analyzing samples with very low RNA input. However, as has been observed with all PCR assays, the Biomark HD demonstrates superior performance on RNA isolated from snap-frozen rather than formalin-fixed tissue samples.

The below referenced articles provide examples of high-throughput qPCR analysis of formalin-fixed tumor biopsy samples using Biomark HD as part of biomarker assessments in clinical trials of anti-PD-L1 (atezolizumab).

• Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients.87

• Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial.33

Conclusions

Although RNA-seq has been increasingly used in biomedical research and may be advantageous in biomarker discovery as an agnostic hypothesis-generating tool, under certain circumstances the aforementioned gene expression platforms may provide significant value, particularly in situations when limited amounts of tissue material or formalin-fixed tissue samples are available. While exon junction (Clariom D) arrays and nCounter assays allow broad interrogation of transcriptomic changes, QGP and high-throughput-qPCR assays are more fit for hypothesis testing or interrogation of specific transcriptomic biomarkers or signatures.

ELISPOT

ELISPOT allows the detection of functionally active, antigen-specific immune cells on the single cell level by capturing the released analyte of interest on a membrane, which is then made visible for enumeration. The assay format as used today for the detection of cytokine-secreting cells was first described in 1988.

• Reverse ELISPOT assay for clonal analysis of cytokine production. I. Enumeration of gamma interferon-secreting cells.314

As a generally easy-to-perform assay with exceptional sensitivity, ELISPOT has remained a common assay choice for basic, translational, and clinical applications in a variety of fields, with surprisingly little change to the basic procedure. A comprehensive review of the technique can be found here.

• Elispot for rookies (and experts too), techniques in life science and biomedicine for the non-expert.315

ELISPOT has been the subject of broad harmonization efforts.

• Results and harmonization guidelines from two large-scale international Elispot proficiency panels conducted by the Cancer Vaccine Consortium (CVC/SVI).316

• Guidelines for the automated evaluation of Elispot assays.317

With the introduction of fluorophores for spot detection (FluoroSpot), multiplexing is now possible and is currently being used for the polyfunctional assessment of cells for up to four different cytokines, resulting in the potential detection of 15 subpopulations. The application of peptide-tagged antigen and anti-tag detection antibodies can be used to identify multiple antibody-secreting cells with different antigen specificities.

• An antigen-specific, four-color, B-cell FluoroSpot assay utilizing tagged antigens for detection.318

While the analysis of ELISPOT and FluoroSpot has largely depended on image analyzers, a new analysis algorithm uses data directly from the camera chip, avoiding any evaluation bias and providing spot volume data, a relative measure of the amount of cytokine released by a single cell.

• Cell detection by functional inverse diffusion and non-negative group sparsity—part I-II.319

An ELISPOT-specific statistical test for response definition has been developed, and an online tool is available to the community for free.

• http://www.scharp.org/zoe/runDFR/ 320

Many publications exist using ELISPOT for immune monitoring purposes in immunotherapeutic cancer trials, including the neoantigen arena.

• An immunogenic personal neoantigen vaccine for patients with melanoma.321

The monitoring of antigen-specific immune responses in patients undergoing immunotherapy can be performed using ex vivo isolated T cells without the bias of in vitro stimulation.

• Immunological markers and clinical outcome of advanced melanoma patients receiving ipilimumab plus fotemustine in the NIBIT-M1 study.322

ELISA

The Enzyme Linked Immunosorbent Assay (ELISA) is a method first described by Weiland323 in 1978 to detect and quantify the presence of analytes, including antibodies, antigens, proteins, and glycoproteins in biological samples.

• The enzyme-linked immunosorbent assay (ELISA)--a new serodiagnostic method for the detection of parasitic infections.323

This method has been widely used for routine diagnosis of viral diseases, such as HIV or HBV infection, but also for pregnancy tests and quantification of soluble molecules in patients’ serum, plasma, urine, or cellular supernatants. ELISAs can be performed in 48-well, 96-well, and 384-well plates, allowing concomitant interrogation of multiple samples, as well as monitoring of changes in analyte concentration at different timepoints.

The most common types of ELISA are the (1) indirect and (2) sandwich methods. The indirect ELISA is generally used to detect antibodies in the serum, plasma, or supernatants. Cancer patients’ humoral immune responses against specific tumor-associated antigens (TAAs) have been assessed with the aim of monitoring changes in immune responses over the course of treatments and to find possible associations with clinical outcome. Humoral responses against shared TAAs, such as NY-ESO-1, p53, and SOX2, have been monitored in serum or plasma of patients with cancer using ELISAs.

• Autoantibodies against cancer antigens.324

• Integrated NY-ESO-1 antibody and CD8⁺ T cell responses correlate with clinical benefit in advanced melanoma patients treated with ipilimumab.325

• Ipilimumab increases activated T cells and enhances humoral immunity in patients with advanced melanoma.326

• Mechanistic insight into the TH1-biased immune response to recombinant subunit vaccines delivered by probiotic bacteria-derived outer membrane vesicles.327

The sandwich method allows the detection of soluble antigens. In this technique, an analyte-specific antibody is coated on the microtiter well. The sample to be analyzed is then added to the well, forming an antigen-antibody complex. A second enzyme-conjugated antibody specific for a different epitope on the antigen is added and, in the presence of an enzyme-specific substrate, the colorimetric reaction is developed. This method allows detection of soluble factors in serum or plasma, including cytokines, immunomodulating molecules, and growth factors, as well as their changes in association with the clinical outcome of patients with cancer undergoing immunotherapy treatments.

• Contribution of humoral immune responses to the antitumor effects mediated by anthracyclines.328

• Soluble NKG2D ligands are biomarkers associated with the clinical outcome to immune checkpoint blockade therapy of metastatic melanoma patients.329

• Immunological markers and clinical outcome of advanced melanoma patients receiving ipilimumab plus fotemustine in the NIBIT-M1 study.322

• A pilot phase I study combining peptide-based vaccination and NGR-hTNF vessel targeting therapy in metastatic melanoma.330

The sensitivity of ELISA can be augmented through the amplified luminescent proximity homogeneous assay (ALPHA). This chemiluminescence-based method can detect analytes at the level of femtograms, reduces washing steps, and is based on the usage of acceptor beads coated with the primary antibody specific for the defined antigen, streptavidin-coated donor beads, and the secondary antibody conjugated to streptavidin.

Multiplexed immunoassays

In the late 1990s, the recently founded Luminex Corporation proposed a commercial FlowMetrix platform for the simultaneous detection of up to 64 analytes using a conventional flow cytometer. The technology relied on 64 distinct sets of fluorescent beads, each coupled with either an antigen, or an antibody, or a nucleic acid, and used as the solid phase of immunoassays or hybridization. The mixing of these distinct sets allowed for the simultaneous monitoring of independent assays using a flow cytometer equipped with digital signal processing.

• Advanced multiplexed analysis with the FlowMetrix system.331

This seminal publication laid the basis of the future development of high dimension protein and nucleic acid profiling platforms. In its current implementation, the Luminex platform is able to monitor 500 simultaneous assays. The versatility of the platform has given rise to several applications in cancer research. A comprehensive review of the origin and evolution of multiplex assays has recently been discussed by Graham H et al. 332

• The genesis and evolution of bead-based multiplexing.322

The simplest application consists of the evaluation of cytokine and chemokine profiles in the serum or plasma of patients affected by cancers.

• The plasma levels of 12 cytokines and growth factors in patients with gastric cancer.333

• Cytokine comparisons between women with breast cancer and women with a negative breast biopsy.334

Luminex assays or other multiplexed bead array embodiments, such as the Cytokine Bead Array (Becton Dickinson), performed on peripheral blood or other body fluids, have been used as prognostic/diagnostic tools in several cancers.

High-parameter flow cytometry

The number of parameters that can be measured in parallel by multicolor flow cytometry has grown rapidly in the last 8 years, mostly due to the development of new polymer-based dyes with tunable emission peaks, as described in the following.

• Brilliant violet fluorophores: a new class of ultrabright fluorescent compounds for immunofluorescence experiments.354

At the same time, new instrumentation has now made it possible to perform flow cytometry with over 18 colors (the previous limit), the limiting factor currently being the fluorochromes. This has allowed the simultaneous monitoring of a large number of immune checkpoint molecules on a wide array of immune cell types, as shown with this 28-color panel.

• OMIP-050: A 28-color/30-parameter fluorescence flow cytometry panel to enumerate and characterize cells expressing a wide array of immune checkpoint molecules.355

Another notable development in the field is flow cytometers that use spectral deconvolution rather than a single bandpass filter/detector for each fluorochrome. In effect, spectral cytometry uses the shape of the complete emission spectrum to determine the signals derived from each fluorochrome on each cell. This has allowed resolution of fluorochromes that would otherwise be very difficult to distinguish. A review of how this method has been applied to cells derived from solid tissues, a common application for tumor immune monitoring, is given in the following.

• Spectral cytometry has unique properties allowing multicolor analysis of cell suspensions isolated from solid tissues.356

The use of standardized multicolor immune monitoring panels to profile immune cell phenotypes and functions in blood and tumor is exemplified in this recent publication in breast cancer.

• Examining peripheral and tumor cellular immunome in patients with cancer.357

Mass cytometry

Flow cytometry is performed using fluorescently tagged antibodies and other fluorescent probes. More recently, the use of heavy metal ion tags, which are chelated to a polymer backbone and read out using time-of-flight mass spectrometry, has been introduced. This method is known as mass cytometry, or CyTOF (Cytometry by Time of Flight). While offering a significantly slower throughput than fluorescence cytometry, it provides the dual advantages of allowing for more antibody specificities to be used in parallel, along with much reduced spillover between detector channels. The method was first described in the following papers.

• Development of analytical methods for multiplex bio-assay with inductively coupled plasma mass spectrometry.358

• Flow cytometer with mass spectrometer detection for massively multiplexed single-cell biomarker assay.359

CyTOF technology was first applied to the study of the immune system in 2010, using a panel of 32 antibodies on human bone marrow cells.

• Single-cell mass cytometry of differential immune and drug responses across a human hematopoietic continuum.360

The number of available mass tags has since grown to over 40, with many publications in different disease settings. It has been used to study signaling in leukemic cells, as well as immune signatures in PBMCs in the context of immunotherapy, as in the following representative publications.

• Mass cytometric functional profiling of acute myeloid leukemia defines cell-cycle and immunophenotypic properties that correlate with known responses to therapy.361

• Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis.362

• Distinct predictive biomarker candidates for response to anti-CTLA-4 and anti-PD-1 immunotherapy in melanoma patients.54

Mass cytometry has also been applied to the study of immune cells infiltrating tumor tissue, as in the following recent publications.

• Interlesional diversity of T cell receptors in melanoma with immune checkpoints enriched in tissue-resident memory T cells.363

• An immune atlas of clear cell renal cell carcinoma.364

A review of technology and important concepts for CyTOF data analysis has been recently published.

• The anatomy of single cell mass cytometry data.365

The above review also links to a website containing a tutorial on mass cytometry data analysis and various tools for the same.

• http://cytof.biosurf.org/ 366

A recent study has made available an ‘Antibody Staining Data Set’ which shows the estimated expression level of 326 cell-surface proteins on 28 different immune cell types, on fresh or fixed cells. The interactive heat map tool can be viewed at the following.

• https://app.astrolabediagnostics.com/antibody_staining_data_set%23:%7E:text=The_Antibody_Staining_Data_Set,subsets_at_single-cell_resolution 367

Multimers

Multimeric forms of MHC–peptide complexes can be identified to stain T cells specific for a given MHC–peptide combination. This was first demonstrated using tetrameric forms of MHC–peptide, bound together using fluorescently labeled streptavidin.

• Phenotypic analysis of antigen-specific T lymphocytes.368

From this work, the term ‘tetramer’ emerged as a common name for these reagents. However, other versions of peptide–MHC multimers have also been made, with backbones other than streptavidin, and with differing resulting valencies. The timeline of development of various MHC multimer reagents is reviewed here.

• Interrogating the repertoire: broadening the scope of peptide-MHC multimer analysis.369

More recent efforts have focused on the parallel use of many different MHC–peptide multimers, which is facilitated by combinatorial staining of each multimer species using a different ‘barcode’ or combination of fluorescent labels.

• Simultaneous detection of many T-cell specificities using combinatorial tetramer staining.370

• Parallel detection of antigen-specific T-cell responses by multidimensional encoding of MHC multimers.371

Such parallelization has also been accomplished using heavy metal-tagged tetramers and mass cytometry.

• Combinatorial tetramer staining and mass cytometry analysis facilitate T-cell epitope mapping and characterization.372

• Multiplexed peptide-MHC tetramer staining with mass cytometry.373

Application of barcoded tetramers, along with nanoparticle display to increase avidity, has been applied to the tracking of neoantigen-specific T cells in tumors and blood.

• Sensitive detection and analysis of neoantigen-specific T cell populations from tumors and blood.374

To create many different MHC–peptide multimers for a given MHC protein, one can use a peptide-exchange approach. With this method, a single MHC protein is produced with an invariant peptide in its binding groove. This is then exchanged for any peptide of interest, as described here.

• Design and use of conditional MHC class I ligands.375

Arrays of MHC–peptide multimers can be used to screen heterogeneous T cell populations for binding. This provides a way to determine the specificities of T cells for neoantigens, for example.

• High throughput determination of the antigen specificities of T cell receptors in single cells.376

Phospho-flow

The use of antibodies to specific phosphoepitopes, in combination with flow cytometry phenotyping, was pioneered in the lab of Garry Nolan, and has been called ‘phospho-flow’. The earliest papers from the Nolan lab described the use of this technique to profile intracellular signaling in either immune or cancer cells.

• Intracellular phospho-protein staining techniques for flow cytometry: monitoring single cell signaling events.377

• Single cell profiling of potentiated phospho-protein networks in cancer cells.378

• Multiparameter analysis of intracellular phosphoepitopes in immunophenotyped cell populations by flow cytometry.379

AbSeq/CITE-seq

In order to analyze the binding of many antibody specificities in parallel on single cells, high-throughput sequencing-based approaches have been developed. These methods use nucleic acid-tagged antibodies and barcoding of single cells with indexing beads, which capture the nucleic acid tags from each cell for high-throughput sequencing. This approach, dubbed AbSeq (antibody-based sequencing), was first published in 2017.

• Abseq: Ultrahigh-throughput single cell protein profiling with droplet microfluidic barcoding.380

It is possible to combine AbSeq with transcriptomic profiling, a method termed cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) or RNA expression and protein sequencing (REAP-seq).

• Simultaneous epitope and transcriptome measurement in single cells.381

• Multiplexed quantification of proteins and transcripts in single cells.382

A protocol for performing such combined proteomic and transcriptomic profiling on a microchip-based single-cell genomics platform has been recently published.

• Protein- and sequencing-based massively parallel single-cell approaches to gene expression profiling.383

Real-time functional assays

The xCELLigence technology (Acea Biosciences) has been developed for cell-based electric impedance assays to monitor and quantify cell proliferation, toxicity, and morphology changes. This platform allows the monitoring of changes in the behavior and adhesion properties of cells through the measurement of impedance, using gold-plated base biosensors in the bottom of culture wells. Cellular status is monitored by label-free and real-time automated reading.

This application can assess cell viability perturbations and determine either cell toxicity and cell death or modification of cell proliferation on treatment with compounds or drugs. Its real-time acquisition of data is suitable to identify the optimal timepoints for a defined cellular activity and to understand the related mechanisms. The xCELLigence technology has increasingly been applied to immunotherapy in order to better understand the complex interaction of immune cells with tumor cells and to verify the efficacy of different immunotherapy approaches or their combinations. Examples are the treatment of cells with either biological agents (eg, agonist or antagonist monoclonal antibodies, or bispecific T cell engagers) or through coculture with effector cells (eg, T, NK, or CAR T cells).

• xCELLigence system for real-time label-free monitoring of growth and viability of cell lines from hematological malignancies.384

• Application of real-time cell electronic analysis system in modern pharmaceutical evaluation and analysis.385

• In vitro immunotherapy potency assays using real-time cell analysis.386

• Avidity characterization of genetically engineered T-cells with novel and established approaches.387

Minimum residual disease detection

Highly sensitive, standardized techniques are necessary for the detection of minimal residual disease (MRD) in order to assess therapy responses and provide more accurate prognoses, leading to improved personalized treatments. Molecular techniques, such as allele-specific oligonucleotide quantitative PCR (ASO-qPCR), represent the traditional approaches.

• High applicability of ASO-RQPCR for detection of minimal residual disease in multiple myeloma by entirely patient-specific primers/probes.388

A major drawback of ASO-qPCR, namely the necessity of using patient-specific probes for B cell malignancies exhibiting somatic hypermutation (eg, multiple myeloma (MM)), is overcome by NGS (also known as deep sequencing or high-throughput sequencing) of immunoglobulin genes.

• Prognostic value of deep sequencing method for minimal residual disease detection in multiple myeloma.389

• A clinical perspective on immunoglobulin heavy chain clonal heterogeneity in B cell acute lymphoblastic leukemia.390

Flow cytometry classically provided lower sensitivity than molecular techniques. However, next-generation flow cytometry (NGF) approaches reach sensitivities comparable with molecular methods, and have the advantage of faster turnaround times, broader applicability, and lower cost.

• Next generation flow for highly sensitive and standardized detection of minimal residual disease in multiple myeloma.391

• ClonoSEQ assay for the detection of lymphoid malignancies.392

While MRD detection is an established tool in the management of hematological malignancies, highly sensitive liquid biopsy assays for the assessment of disseminated tumor cells in the context of solid tumors have been developed more recently. These probe for CTCs or circulating tumor-derived factors such as ctDNA, as for example, Natera’s Signatera assay, which provides early disease recurrence prediction based on ctDNA. Since CTCs are rare events, marker-dependent (such as Menarini’s DEPArray) or marker-independent techniques (eg, microfiltration) are applied to enrich them prior to downstream analysis.

• Analysis of plasma cell-free DNA by ultradeep sequencing in patients with stages I to III colorectal cancer.393

• Liquid biopsy and minimal residual disease – latest advances and implications for cure.394

These techniques as well as additional techniques such as droplet digital PCR (ddPCR), positron emission tomography (PET)-CT, PET-MRI, CyTOF, and monoclonal immunoglobulin rapid accurate mass measurement (miRAMM) are being explored as complementary or alternative approaches for MRD assessment as discussed here.

• Droplet digital PCR for minimal residual disease detection in mature lymphoproliferative disorders.395

• Minimal residual disease in multiple myeloma: impact on response assessment, prognosis and tumor heterogeneity.396

Neoantigens

Neoantigens are potentially immunogenic epitopes that are created by tumor mutations or chromosomal rearrangements in the course of an individual’s cancer development. They have emerged as an important target of antitumor T cells, and this reactivity may be important both in the context of ICIs as well as adoptive cell therapy (ACT).

Proliferation and cytotoxicity assays

Ex vivo or in vitro cell proliferation can be determined by multiple technologies. The earliest approach is to label cells with [³H]-thymidine and quantify proliferation by a gamma counter, which corresponds to DNA synthesis, and has been used to monitor ex vivo immune responses to cancer vaccines. Bromodeoxyuridine (BrdU) incorporation assays use absorbance as the readout and avoid radiolabeled dyes. Other cell proliferation assays use dyes that measure the metabolic activity of cells (eg, MTT (using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide), WST-1 (using water-soluble tetrazolium salts)) by permeabilizing the cells and reacting with enzymes or other metabolic factors.

• Assays for monitoring cellular immune responses to active immunotherapy of cancer.413

• Immunotherapy of metastatic malignant melanoma by a vaccine consisting of autologous interleukin-2-transfected cancer cells: outcome of a Phase I study.414

• Generation of immunity to the HER-2/neu oncogenic protein in patients with breast and ovarian cancer using a peptide-based vaccine.415

Fluorescent dyes can be used to monitor cell populations with different rates of division and proliferation. Carboxyfluorescein succinimidyl ester (CFSE) has been commonly used to assess lymphocyte division and proliferation by flow cytometry. CFSE can covalently label intracellular molecules, and CFSE concentration is proportionally diluted in cell progeny according to the number of subsequent cell divisions. DELFIA (dissociation-enhanced lanthanide fluorescence immunoassay) is based on the measurement of time-resolved fluorescence intensity. These fluorescence-based dyes are compatible with multiparametric staining with other fluorochromes.

• The use of carboxyfluorescein diacetate succinimidyl ester (CFSE) to monitor lymphocyte proliferation.416

• Role of STAT3 in CD4⁺ CD25⁺ FOXP3⁺ regulatory lymphocyte generation: implications in graft-versus-host disease and antitumor immunity.417

The efficiency of in vitro expansion of T cells engineered to express exogenous antigen-specific CAR or TCR measured by proliferation assays can represent a relevant parameter and a surrogate marker of possible in vivo persistence.

• Adoptive cell transfer therapy following non-myeloablative but lymphodepleting chemotherapy for the treatment of patients with refractory metastatic melanoma.418

• CAR T cell immunotherapy for human cancer.419

• CARs on track in the clinic.420

Cytotoxicity assays measure cell death induced by cytotoxic stimuli, environmental changes, or cell-mediated killing. They are based on cell membrane integrity, using vital dyes that allow the exclusion of viable cells, or assessing the release of markers from dying cells (eg, CFSE). In addition, metabolic activity measurements, for example, MTT, lactate dehydrogenase (LDH), or ATP assays, are also used to measure cell viability. A reliable cell-mediated cytotoxicity assay based on radiolabeled ⁵¹Cr release has been substituted by the development of the assays mentioned above. DELFIA also represents a reliable and simple technology to asses either antibody-dependent cell-mediated cytotoxicity (ADCC) or cell-mediated cytotoxic assays. These assays have been applied to the functional characterization of antitumor and antigen-specific T and NK cell responses, highlighting relevant therapeutic implications.

• Optimization of cytotoxic assay by target cell retention of the fluorescent dye carboxyfluorescein diacetate (CFDA) and comparison with conventional 51CR release assay.421

• WT1 peptide-specific T cells generated from peripheral blood of healthy donors: possible implications for adoptive immunotherapy after allogeneic stem cell transplantation.422

• Preparation of cytokine-activated NK cells for use in adoptive cell therapy in cancer patients: protocol optimization and therapeutic potential.423

Assessment of ex vivo antigen-specific immune responses

TAAs are recognized by T cells in the form of MHC/peptide complexes. The discovery and molecular characterization of TAAs allowed the identification and categorization of three main types of antigens.

1. Differentiation antigens that are specific to the cellular lineage. These antigens are overexpressed by tumor cells but also shared with normal tissue (eg, carcinoembryonic antigen (CEA), epithelial cell adhesion molecule (Ep-CAM), mucin-1 (MUC-1), melanoma antigen recognized by T cells-1 (MART-1/Melan-A), glycoprotein 100 (gp100), and tyrosinase (Tyr)).

2. Cancer-Testis (CT) antigens (eg, MAGE, GAGE, LAGE, and NY-ESO-1) are expressed in tumors with different histological origins and their expression in normal tissues is limited to testicular germ cells and placenta.

3. Mutated antigens or neoantigens derived from non-synonymous somatic mutations. These TAAs are not expressed by normal cells and display superior immunogenic potency as compared with differentiation/self or CT antigens. Details of these TAAs have been provided above.

Single or multiple peptides derived from the TAAs mentioned above have been administered in the context of phase I/II clinical studies for vaccination of patients with different types of tumors. However, cancer vaccines have not been associated with durable clinical responses in patients with cancer, possibly as a result of low immunogenic potency and the pre-existence in patients’ body of tolerogenic/anergic T cells specific for these TAAs. Nevertheless, antigen-specific T cell responses could be detected in the circulation of up to 50% of patients with cancer undergoing vaccination. Of note, antigen-specific T cell responses in the circulation of patients with cancer could be detected ex vivo with, in some cases, association with their clinical outcome.

• Progress in the development of immunotherapy for the treatment of patients with cancer424

• Cancer/testis antigens, gametogenesis and cancer.425

• Therapeutic vaccines for cancer: an overview of clinical trials.426

The monitoring of circulating antigen-specific T cell responses in patients with cancer can be performed by ex vivo coculture of peripheral blood lymphocytes with either peptides containing TAA-derived epitopes or HLA-matched antigen-presenting cells loaded with peptides. Cytokine release assays, such as ELISA, ELISPOT, or intracellular staining by flow cytometry, have been used as readouts of antigen-specific reactivity by T cells. These methods have been described above. The majority of studies assessed the effector functions of antigen-specific T cells by measuring the release of IFN-γ, but in some cases also analyzed the multifunctional properties of T cells by assessing the release of multiple cytokines. Multimer staining in combination with flow cytometry has been used to determine the antigen-specific reactivity of T cells.

• Flow cytometric determination of intracellular or secreted IFNgamma for the quantification of antigen reactive T cells.427

• Immunologic monitoring of cancer vaccine trials using the ELISPOT assay.428

• MHC class II tetramer analyses in AE37-vaccinated prostate cancer patients reveal vaccine-specific polyfunctional and long-lasting CD4⁺ T-cells.429

• Anti-CTLA-4 antibody therapy: immune monitoring during clinical development of a novel immunotherapy.430

Cancer vaccine studies based on the usage of neoantigens as a source of immunization also showed both the expansion of pre-existing T cell responses and the induction of new T cells reactive against this type of antigen in patients with objective clinical responses.

The ex vivo assessment of antigen-specific reactivity, including neoantigens, could also be successfully performed for TILs, and represented the rationale for ACT protocols for patients with solid tumors (eg, melanoma, CRC, and breast cancer). The efficient recognition by TILs of mutated antigens expressed by autologous tumor cells has been associated with tumor regressions following ACT.

• Somatically mutated tumor antigens in the quest for a more efficacious patient-oriented immunotherapy of cancer.431

• Exploiting the curative potential of adoptive T-cell therapy for cancer.432

• Tumor exome analysis reveals neoantigen-specific T-cell reactivity in an ipilimumab-responsive melanoma.120

Multiplex immunofluorescent staining

Tissue IHC, a century-old technology, has undergone a major technological revolution in recent years. With the advancement of three major components, (1) biomarker staining methods, (2) whole slide imaging (WSI) techniques, and (3) image analysis software, multiplexing technologies are slowly replacing conventional single-plex IHC assays. Tissue multiplexing technologies play a vital role in understanding the complex TME in cancer immunotherapy. Staining multiple protein biomarkers on a single tissue section facilitates an understanding of complex cell–cell interactions, cell migration and infiltration of immune cells, and cellular distance and density.

Staining methods

Three different staining methods are gaining popularity: (1) chromogenic multiplexing, (2) fluorescent multiplexing, and (3) mass spectrometry.

Fluorescent multiplexing

Multiplex immunofluorescent staining technology allows multiple biomarkers tagged with distinct fluorescent dyes to be interrogated separately or in combination, and has exhibited exponential growth in recent years. Although fluorescent multiplex assays are predominantly research use only or lab-derived test, future clinical adoption is anticipated. The technology requires investment in fluorescent microscopes and scanners, and relies heavily on image analysis software for downstream analysis and effective outcomes.

This section highlights three different fluorescent staining methodologies:

Mass spectrometry for tissue multiplexing

Mass spectrometry is used for identifying multiple biomarkers from a single tissue section. Typically, the tissue section is incubated with a cocktail of antibodies tagged with unique metal tags, each targeting a different protein of interest. Such methodologies are supported by dedicated imaging systems and image analysis software. Two technologies available in the market are highlighted here.

Whole slide imaging and image analysis

Application of tissue multiplexing technology

Different groups have successfully applied the complex multiplexing workflow to answer key questions in the field of cancer immunotherapy. The following are some of the recent publications.

• Systems pathology by multiplexed immunohistochemistry and whole-slide digital image analysis.448

• Validation of multiplex immunofluorescence panels using multispectral microscopy for immune-profiling of formalin-fixed and paraffin-embedded human tumor tissues.449

• Multispectral fluorescence imaging allows for distinctive topographic assessment and subclassification of tumor-infiltrating and surrounding immune cells.450

• Multiplex immunohistochemistry for molecular and immune profiling in lung cancer-just about ready for prime time?451

A comprehensive review of complex tissue multiplexing technology is provided here.

• State-of-the-art of profiling immune contexture in the era of multiplexed staining and digital analysis to study paraffin tumor tissues.452

Gene expression analysis tools

Data analyses of gene expression profiling can be categorized depending on the study objectives, that is, class discovery, class comparison, class prediction, and survival analysis.

• Design and analysis of DNA microarray investigations.453

Class discovery is an unsupervised method with the goal of discovering clusters among specimens or among genes. Hierarchical clustering algorithms, principal component analysis, a self-organizing map, and non-negative matrix factorization are popular methods to find clusters of samples or genes.

• Cluster analysis and display of genome-wide expression patterns.454

• Principal components analysis to summarize microarray experiments: application to sporulation time series.455

• Analysis of gene expression data using self-organizing maps.456

• Knowledge-based gene expression classification via matrix factorization.457

Class comparison aims to identify genes differentially expressed between two or more different tissue types or experimental conditions. For microarray gene expression data, a Student’s t-test is generally used to identify significantly differentially expressed genes between two phenotype classes, assuming the log-transformed ratio or intensity data are approximately normally distributed. A multiple testing problem occurs during class comparison due to the simultaneous testing of all genes. The false discovery rate method is commonly used to control for false discoveries in a set of identified genes.

• Microarrays, empirical Bayes methods, and false discovery rates.458

• Significance analysis of microarrays applied to transcriptional responses to ionizing radiation.459

• Controlling the false discovery rate: a practical and powerful approach to multiple testing.460

Class prediction is used to predict the phenotype of new samples from their gene expression. Popular prediction methods include diagonal linear discriminant analysis, K-nearest neighbors, support vector machine, and random forest. Lasso and least angle regression (LARS) can be used to build prediction models with continuous outcomes. Cross validation needs to be properly performed to avoid overfitting, and the final model always needs to be evaluated on independent data sets.

• Regularized linear discriminant analysis and its application in microarrays.461

• Diagnosis of multiple cancer types by shrunken centroids of gene expression.462

• Gene selection for sample classification based on gene expression data: study of sensitivity to choice of parameters of the GA/KNN method.463

• Knowledge-based analysis of microarray gene expression data by using support vector machines.464

• Gene selection and classification of microarray data using random forest.465

• The elements of statistical learning: data mining, inference, and prediction (Springer Series in Statistics).466

• The lasso method for variable selection in the Cox model.467

• A paradigm for class prediction using gene expression profiles.468

Gene expression-based survival analyses address two major areas: (1) identifying signatures that are associated with survival, and (2) building a prognostic and/or predictive model based on survival data. One method uses a Cox proportional hazards model to relate survival time to k ‘supergene’ expression groups. The ‘supergene’ expression levels are the first k principal component, that is, linear combinations of expression levels of the subset of genes that are univariately correlated with survival. Another method uses a penalized Cox regression model to find genes related to survival, which can be done using the R package glmnet.

• Diagnostic and prognostic prediction using gene expression profiles in high-dimensional microarray data.469

• ‘Gene shaving’ as a method for identifying distinct sets of genes with similar expression patterns.470

• Regularization paths for Cox’s proportional hazards model via coordinate descent.471

The R statistical programming language together with the Bioconductor repository provides over 1000 packages for the analysis of high-throughput genomic data including gene expression. For users with limited or no programming skills, a more suitable choice may be GEO2R, which can provide simple data analysis on a curated National Center for Biotechnology Information (NCBI) GEO data set. Such users can also use a general-purpose software that can handle more sophisticated data analyses. For example, BRB-ArrayTools is a Windows desktop application with a graphical interface designed for use by researchers who want to use state-of-the-art statistical methods for gene expression analysis.

• Bioinformatics and computational biology solutions using R and Bioconductor.472

• GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor.473

• Analysis of gene expression data using BRB-ArrayTools.474

WES and RNA-seq data analysis tools

The first step in the workflow of WES/RNA-seq data is mapping. Raw sequence data are commonly saved in the FASTQ format, which needs to be mapped to a reference genome by using an alignment algorithm. Popular mapping/alignment tools include Burrows-Wheeler Aligner (BWA)-Maximal Exact Match (MEM) and Subread for DNA-seq data, and TopHat2 and spliced transcripts alignment to a reference (STAR) for RNA-seq data.

• TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions.475

• STAR: ultrafast universal RNA-seq aligner.476

• The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote.477

• Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM.478

Next, variant calling can be conducted on both aligned RNA-seq and DNA-seq data using Samtools or Genome Analysis Toolkit best practices pipeline tools. For WES data with tumor/normal pairs, somatic mutations are called using tools such as MuTect (for SNPs) and Mutect2 (for indels). The variant call format files containing variant information are finally annotated using tools such as ANNOVAR, SnpEff, and variant effect predictor (VEP).

• From FastQ data to high confidence variant calls: the Genome Analysis Toolkit best practices pipeline.479

• The sequence alignment/map format and SAMtools.480

• Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples.481

• ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data.482

• A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3.483

• The Ensembl variant effect predictor.484

Further levels of analyses based on WES data include using tools such as Sequenza to generate allele-specific copy numbers from the mapped reads and PyClone to generate the tumor clonal distribution. To get an estimate of tumor purity and ploidy, fraction and allele-specific copy number estimates from tumor sequencing (FACETS) can be applied on mapped reads. In addition, PolySolver can be used to predict patient HLA types, and the MSI status can be detected using a method called microsatellite instability by NGS (mSINGS). Finally, multiple pipelines have been developed to predict neoantigens based on annotated variants in the translated protein sequence.

• Sequenza: allele-specific copy number and mutation profiles from tumor sequencing data.485

• PyClone: statistical inference of clonal population structure in cancer.486

• FACETS: allele-specific copy number and clonal heterogeneity analysis tool for high-throughput DNA sequencing.487

• Comprehensive analysis of cancer-associated somatic mutations in class I HLA genes.488

• Microsatellite instability detection by next generation sequencing.489

• Applications of immunogenomics to cancer.490

To generate quantitative measurements from the aligned RNA-sequence data, the high-throughput sequencing (HTSeq) method and featureCounts software can be applied to generate count data. Similar tools include RNA-seq by expectation maximization (RSEM) and Salmon. Due to the nature of RNA-seq data, differential expression analysis for sequence count data (DESeq), DESeq2, Cuffdiff, and EdgeR use a negative binomial distribution with generalized linear models to determine the significance of genes that are differentially expressed between classes. The mapped reads from RNA-seq data can be further explored to study the immune repertoire of the TCR and B cell receptor using tools such as TCR repertoire utilities for solid tissue/tumor (TRUST), and to analyze immune cell composition using tools such as cell type identification by estimating relative subsets of RNA transcripts (CIBERSORT) and the tumor immune estimation resource (TIMER).

• HTSeq-a Python framework to work with high-throughput sequencing data.491

• featureCounts: an efficient general-purpose program for assigning sequence reads to genomic features.492

• RSEM: accurate transcript quantification from RNA-seq data with or without a reference genome.493

• Salmon provides fast and bias-aware quantification of transcript expression.494

• Differential expression analysis for sequence count data.495

• Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.496

• edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics.497

• Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks.498

• Antigen receptor repertoire profiling from RNA-seq data.499

• Profiling tumor infiltrating immune cells with CIBERSORT.500

• Comprehensive analyses of tumor immunity: implications for cancer immunotherapy.501

For scRNA-seq data, scalability and technical noise are two main bioinformatics challenges. Certain methods that were developed for bulk cell RNA-seq cannot be applied directly and require adaptation. Many methods and tools have been developed specifically for scRNA-seq data. For example, Scater, SCnorm, Seurat, and SCANPY are R or Python packages developed to facilitate rigorous preprocessing, quality control, normalization, and visualization of scRNA-seq data. Depending on the experimental design, protocol, and platform, SCONE uses a data-driven approach to allow the user to select an appropriate normalization strategy for scRNA-seq data. Falco is a cloud-based scRNA-seq processing framework which provides a scalable and efficient computational solution. For data visualization, many tools, such as t-distributed stochastic neighbor embedding (t-SNE), uniform manifold approximation and projection (UMAP), and single-cell interpretation via multi-kernel learning (SIMLR) can be used as a dimensionality reduction step for visualizing scRNA-seq data in two dimensions. A typical scRNA-seq analysis by Luecken and Theis502 serves as a strong example of best practices for this technique.

• Computational and analytical challenges in single-cell transcriptomics.291

• Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R.503

• SCnorm: robust normalization of single-cell RNA-seq data.504

• Integrating single-cell transcriptomic data across different conditions, technologies, and species.505

• SCANPY: large-scale single-cell gene expression data analysis.506

• Performance assessment and selection of normalization procedures for single-cell RNA-seq.507

• Falco: a quick and flexible single-cell RNA-seq processing framework on the cloud.508

• Dimensionality reduction for visualizing single-cell data using UMAP.509

• Visualization and analysis of single-cell RNA-seq data by kernel-based similarity learning.510

• Current best practices in single-cell RNA-seq analysis: a tutorial.502

Multiparameter cytometric data analysis

One of the main goals of multiparameter cytometry data analysis is to identify the proportion of immune cell subpopulations (eg, memory CD8⁺ T cells, Th1 cells, and so on) and their properties (eg, expression of activation markers or cytokines, antigen-specificity of T cells, and so on).

Analysis of flow cytometry data starts from data preprocessing (compensation, batch effect assessment and removal, curation of data sets, and exclusion of dead cells and doublets) followed by placing the ‘gates’ to identify cell subsets and obtain property information of those subsets. Postprocessing entails comparison of groups and aims to find significantly altered subpopulations of cells with appropriate statistical methods.

• Guidelines for the use of flow cytometry and cell sorting in immunological studies.511

Traditionally, flow cytometric data are manually analyzed using a dedicated software with a graphical user interface, for example:

• https://www.flowjo.com/ 512

• https://www.denovosoftware.com/ 513

• https://www.aceabio.com/products/novoexpress-software/ 514

• https://www.vsh.com/products/winlist/index.asp 515

• https://www.miltenyibiotec.com/US-en/products/macs-flow-cytometry/software/flowlogic-software.html 516

• https://www.beckman.com/flow-cytometry/software/kaluza 517

Similar to flow cytometry data analysis, mass cytometry data also require preprocessing (eg, transformation of ion counts, normalization and batch correction, and so on).

• The anatomy of single cell mass cytometry data.365

Recent developments in both multiparameter flow and mass cytometry enable the interrogation of up to 40 parameters at the single cell level. This high dimensionality requires sophisticated computational approaches when analyzing data to gain worthwhile insights from information-rich data sets. As the list of available analysis tools continues to grow, it is worthwhile to start from a high-level overview. There are supervised and unsupervised approaches. We will cover unsupervised approaches first, followed by supervised approaches.

The most commonly used unsupervised analyses are nicely summarized here.

• Algorithmic tools for mining high-dimensional cytometry data.518

• Computational flow cytometry: helping to make sense of high-dimensional immunology data.519

• The end of gating? An introduction to automated analysis of high dimensional cytometry data.520

Unsupervised approaches for multiparametric flow or mass cytometry have a significant overlap with those for single-cell NGS data sets. In general, these can be classified as (1) data visualization, (2) cell type identification, (3) differential analysis, and (4) network and multiomics data integration.

• A beginner’s guide to analyzing and visualizing mass cytometry data.521

• Computational approaches for high-throughput single-cell data analysis.522

• Advancing systems immunology through data-driven statistical analysis.523

• Mass cytometry: a powerful tool for dissecting the immune landscape.524

While most analysis tools are written in open source programing languages such as R, Python, or others, fee-based platform services are available for multiparametric cytometry data.

• https://www.cytobank.org/ 525

• https://www.astrolabediagnostics.com/ 526

Benchmark assessment of algorithms is a helpful resource when deciding what algorithms to use, although it is recommended to use multiple algorithms to ensure the validity of analysis results.

• Critical assessment of automated flow cytometry data analysis techniques.527

• Comparison of clustering methods for high-dimensional single-cell flow and mass cytometry data.528

A supervised approach is also useful if population and biomarker candidates are well defined with the traditional cascaded gating strategy. In this case, automated gating reduces both interanalysts’ variabilities and time spent for manual gating.

• Flow cytometry bioinformatics.529

• Standardizing flow cytometry immunophenotyping analysis from the human immunophenotyping consortium.530

• A standardized immune phenotyping and automated data analysis platform for multicenter biomarker studies.531

• Implementation and validation of an automated flow cytometry analysis pipeline for human immune profiling.532

Software and tools for function and pathway analysis

Functional enrichment and pathway analysis are essential tasks for the interpretation of gene lists derived from large-scale genetic, transcriptomic and proteomic studies. Two of the most popular resources for understanding high-level functions and utilities of the biological system are the Kyoto Encyclopedia of Genes and Genomes (KEGG) and BioCarta. A comprehensive list of available tools is provided at the Gene Ontology website. ImmuneSigDB is a comprehensive compendium of 5000 gene sets pertaining to immune biology, which may provide the systems immunologist with a useful resource for analysis of gene expression in the immune system.

• http://www.geneontology.org 533

• KEGG: Kyoto Encyclopedia of Genes and Genomes.534

• The Gene Ontology (GO) database and informatics resource.535

• Compendium of immune signatures identifies conserved and species-specific biology in response to inflammation.536

Genomic, proteomic, and metabolomic data typically result in lists of interesting genes or proteins. Translating these gene sets into an understanding of the underlying biological mechanism is a fundamental need in biological research. A popular resource that can help unravel the mechanism behind a specific set of mutations is the Database for Annotation, Visualization, and Integrated Discovery (DAVID). Pathway Commons is another tool used by computational biologists to download custom subsets of pathway data for analysis, or to incorporate powerful biological pathway and network information retrieval and query functionality into websites and software. One widely used commercial product is Ingenuity Pathway Analysis, which allows the user to access many different algorithms to identify the most significant pathways and discover potential novel regulatory networks and causal relationships associated with experimental data.

• DAVID: https://david.ncifcrf.gov 537

• Pathway Commons: http://www.pathwaycommons.org/pc2 538

• DAVID bioinformatics resources: expanded annotation database and novel algorithms to better extract biology from large gene lists.539

• Pathway Commons, a web resource for biological pathway data.540

• Causal analysis approaches in Ingenuity Pathway Analysis.541

Gene set enrichment analysis (GSEA) is a computational method that determines whether a specific set of genes shows statistically significant, concordant differences between two biological states. The gene-level statistics for all genes in a pathway are aggregated into a single pathway-level statistic. Finally, statistically significant pathways can be identified. The tools for this kind of analysis are distributed as either stand-alone desktop applications or as packages for R (eg, R-GSEA). Gene set analysis (GSA) is an R function that differs from GSEA in its use of the ‘maxmean’ statistic, which is often more powerful than the modified Kolmogorov-Smirnov statistic used in GSEA. GSA can also handle more than two groups, such as multiple classes, survival times, and quantitative outcomes. Both GSEA and GSA approaches can help identify the most promising pathways or gene sets to be used as predictive or prognostic biomarkers in immunotherapy.

• Ten years of pathway analysis: current approaches and outstanding challenges.542

• Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles.543

• Exploring gene expression data with class scores.544

• On testing the significance of sets of genes.545

Conclusions

The advancement of high-throughput technologies has provided an unprecedented opportunity to conduct comprehensive analyses of genes, transcripts, proteins, and other significant biological molecules for biomarker identification. However, it has also complicated the process of finding meaningful markers from these complex data sets. It is time-consuming and challenging to validate the accuracy of each step involved in the analysis workflow. In order to obtain reproducible results, it is imperative to harmonize not only sample preparation and assay execution protocols but also data analysis procedures.

Many software packages and pipelines have been developed that have allowed us to effectively process WES/RNA-seq data and multiparameter cytometry data, as well as to perform gene set/pathway analyses. However, depending on the methods used in each data processing and analysis step, substantially different results and conclusions may be developed from the same data set. These results should be assessed and validated carefully with different methods and experiments.

Imaging cytotoxic lymphocytes

• Synthesis of 2’-deoxy-2’-[¹⁸F]fluor-9-beta-D-arabinofuranosylguanine: a novel agent for imaging T cell activation with PET.550

• Molecular imaging of lymphoid organs and immune activation by positron emission tomography with a new ¹⁸F-labeled 2’-deoxycytidine analog.551

• Noninvasive detection of therapeutic cytolytic T cells with ¹⁸F-FHBG PET in a patient with glioma.552

Imaging immunosuppressive factors and cells

• TGF-β antibody uptake in recurrent high-grade glioma imaged with ⁸⁹Zr-fresolimumab PET.553

• Clinical applications of iron oxide nanoparticles for magnetic resonance imaging of brain tumors.554

• Antibody positron emission tomography imaging in anticancer drug development.555

Preclinical studies continue to explore molecular imaging probes that enable visualization of immune responses within tumors. Immuno-PET, the use of antibodies or antibody fragments to target PET radionuclides, introduces increased specificity to imaging, with new probes appearing with increasing frequency.

• An effective immuno-PET imaging method to monitor CD8-dependent responses to immunotherapy.556

Antibody engineering optimizes in vivo pharmacokinetics and provides improved blood clearance for imaging at early timepoints with high tissue specificity.

• Engineered antibody fragments for immuno-PET imaging of endogenous CD8⁺ T cells in vivo.557

• Targeting T and B lymphocyte with radiolabeled antibodies for diagnostic and therapeutic applications.558

• Noninvasive imaging of immune responses.559

Immuno-PET probes targeted to cytokines (eg, IL-1β and IFN-γ) show increased tissue specificity over probes targeted at immune cell types that show significant uptake in secondary lymphoid tissues.

• Immuno-PET of innate immune markers CD11b and IL-1β detect inflammation in murine colitis.560

• IFN-γ PET Imaging as a predictive tool for monitoring response to tumor immunotherapy.561

Radiolabeling of anti-PD-L1 antibodies and small non-antibody therapeutics allows for the in vivo distinction of PD-L1 positive and negative tumors.

• Noninvasive imaging of tumor PD-L1 expression using radiolabeled anti-PD-L1 antibodies.562

• Engineering high affinity PD-1 variants for optimized immunotherapy and immuno-PET imaging.563

Given the intensity of work on CAR T cell therapy, significant efforts are underway to generate reporter systems to track the distribution, persistence, and in situ function of transferred T cells.

• [¹⁸F]FHGB PET/CT imaging of CD34-TK75 transduced donor T cells in relapsed allogeneic stem cell transplant patients: safety and feasibility.564

• Quantitative imaging of the T cell antitumor response by positron-emission tomography.565

• Imaging TCR-dependent NFAT-mediated T-cell activation with positron emission tomography in vivo.566

Finally, a growing mechanistic understanding of T cell metabolism in TME enables in vivo quantification of effector T cell accumulation pretherapy and post-therapy through metabolite detection.

• Lactate chemical exchange saturation transfer (LATEST) imaging in vivo a biomarker for LDH activity.567

• Molecular imaging biomarkers for cell-based immunotherapies.568

Conclusions

Molecular imaging allows non-invasive profiling of whole primary tumors, distal metastases, and involved lymph nodes, potentially aiding patient selection for given therapies and evaluation of response. Molecular imaging is just starting to be used to monitor therapy in immuno-oncology, and preclinical models indicate the potential to monitor specific cellular processes in a longitudinal manner independent of biopsy bias.

Tryptophan-degrading enzymes

The metabolic by-products of tryptophan (Trp) degradation including kynurenine (kyn) have been implicated as playing an important role in the generation of immune tolerance. Given that IDO-1 is upregulated in more inflamed environments due to its regulation by IFN signaling, expression levels often correlate with PD-L1 expression and numbers of infiltrating CD8⁺ T cells. As a result, several studies have found positive associations between IDO-1 expression and response to checkpoint blockade.

• A prospective phase II trial exploring the association between tumor microenvironment biomarkers and clinical activity of ipilimumab in advanced melanoma.141

• An immune-active tumor microenvironment favors clinical response to ipilimumab.569

• High IDO-1 expression in tumor endothelial cells is associated with response to immunotherapy in metastatic renal cell carcinoma.570

These data seemingly contradict additional studies that have also linked IDO-1 expression by certain tumor models with resistance to anti-CTLA-4 antibody immunotherapy. Consistent with this finding, IDO-1 can promote the recruitment of myeloid-derived suppressor cells (MDSCs), making these tumors more susceptible to agents blocking MDSC recruitment.

• Indoleamine 2, 3-dioxygenase is a critical resistance mechanism in antitumor T cell immunotherapy targeting CTLA-4.571

• Targeting myeloid-derived suppressor cells with colony stimulating factor-1 receptor blockade can reverse immune resistance to immunotherapy in indoleamine 2,3-dioxgenase-expressing tumors.572

The IDO-1 enzyme converts Trp to a series of metabolic by-products including kyn. While this process starves effector T cells of the essential amino acid Trp and suppresses their proliferation, kyn drives regulatory T cell differentiation and promotes immune tolerance. Indeed, the following reference describes serum kyn/Trp ratios and quinolinic acid concentrations based on high-performance liquid chromatography-mass spectrometry (HPLC-MS) and MS analyses as being associated with diminished PFS in patients with NSCLC undergoing nivolumab (anti-PD-1 antibody) immunotherapy.

• Can IDO activity predict primary resistance to anti-PD-1 treatment in NSCLC?.573

Notably, these data are also consistent with a recent study demonstrating that adaptive increases in serum kyn/Trp ratio are associated with inferior OS in patients with advanced melanoma and RCC undergoing nivolumab immunotherapy.

• Metabolomic adaptations and correlates of survival to immune checkpoint blockade.574

Adenosine

Adenosine is an immunosuppressive nucleoside that has been demonstrated to negatively regulate antitumor immunity by directly suppressing effector T cell and NK cell functions. CD73 is an ecto-5’-nucleotidase that controls levels of adenosine in the TME by catalyzing the breakdown of adenosine monophosphate (AMP). Preclinical studies indicate that tumor expression of CD73 is associated with inferior responses to checkpoint inhibitor immunotherapy and increased responsiveness to the inhibition of adenosine signaling. CD73 can also be cleaved from the cell surface and found in the serum. Recent work has shown that the enzymatic activity of soluble CD73 in the serum correlates with response to nivolumab in patients with metastatic melanoma.

• Antimetastatic effects of blocking PD-1 and the adenosine A2A receptor.575

• Adenosine receptor 2A blockade increases the efficacy of anti-PD-1 through enhanced antitumor T cell responses.576

• Soluble CD73 as biomarker in metastatic melanoma patients treated with nivolumab.577

Tumor glycolysis and oxidative phosphorylation

The metabolic landscape of the TME has been shown to contribute to immunotherapy resistance in a variety of contexts. The elevated glycolytic capacity of malignant tissues has been associated with T cell glucose starvation and diminished antitumor immune responses. Indeed, circulating levels of LDH, which may serve as a surrogate for glycolytic levels in the TME, also correlate with inferior responses to checkpoint inhibitor therapy.

• Increased tumor glycolysis characterizes immune resistance to adoptive T cell therapy.578

• Targeting tumor-associated acidity in cancer immunotherapy.579

Other studies have indicated that the process of oxidative phosphorylation can also be associated with resistance to anti-PD-1 blockade in both melanoma cell lines and clinical melanoma specimens.

• Tumor cell oxidative metabolism as a barrier to PD-1 blockade immunotherapy in melanoma.580

Oxidative phosphorylation of tumors has been associated with the development of tumor hypoxia, a process that has also correlated with resistance to pembrolizumab immunotherapy in preclinical models and patients with advanced melanoma.

• Genomic and transcriptomic features of response to anti-PD-1 therapy in metastatic melanoma.166

• Efficacy of PD-1 blockade is potentiated by metformin-induced reduction of tumor hypoxia.581

Fatty acid oxidation (FAO) drives oxidative phosphorylation and in local myeloid cells within the TME this process may contribute to tumor-mediated immune evasion. Consistent with an important role for FAO in the regulation of tumor immunity, the genetic silencing and systemic pharmacological inhibition of this pathway in preclinical models enhance the efficacy of immunotherapy.

• Inhibition of fatty acid oxidation modulates immunosuppressive functions of myeloid-derived suppressor cells and enhances cancer therapies.582

• Paracrine Wnt5a-β-catenin signaling triggers a metabolic program that drives dendritic cell tolerization.583

Conclusions

In light of the emerging data highlighting the critical role of cellular metabolism in the regulation of antitumor immunity, the role of metabolic biomarkers in the development of novel immunotherapy strategies is likely to expand. The development of methods to more readily study in situ metabolic biomarkers in a cell type-specific manner would greatly facilitate the use of metabolic biomarkers in clinical trial development.