Christelle Johnson, MS, PhD
Senior Field Applications Scientist
Cancer Genomics & Immuno-Oncology

Comprehensive Immunogenomics Profiling for the Discovery of the Next Generation of Oncology Biomarkers

Although hypothesized back in the early 1900s that the immune system may provide a protective mechanism against tumorigenesis, only recently have the fields of immunology and oncology been better linked. It is well accepted today that the immune system protects the host from cancer by eliminating malignant cells. However, tumor cells may gain an ability to evade the immune system by blocking key interactions needed for immune response. Overcoming this barrier has been the general goal of recent immunotherapies and combination agents that can restore immune function. The use of predictive biomarkers has helped researchers identify promising targets for immunotherapy, as well as enrich for patients who are more likely to respond to such therapy.

The term “biomarker” is broad, as it includes any biological molecule that can be detected and measured in human tissues or body fluids. A consensus definition given by a Biomarkers Working Group is ‘an objective, measurable indicator of normal physiology, or pathophysiological process, or response to pharmacological intervention’ (1). When discussing immunotherapies, it is this latter part of the definition that becomes most relevant. Emphasis was placed on the need to separate “clinical endpoints” from “surrogate endpoints” in therapeutic interventions (1). Patient survival and quality of life are considered examples of clinical endpoints. On the other hand, biomarkers serve as surrogate endpoints to evaluate patient response to treatment.

PD-L1, MSI, and TMB in clinical practice

Biomarkers aid in predicting beneficial drug response. A classic example of this strategy is immunohistochemistry (IHC) staining of tumor biopsies to determine PD-L1 expression status. Patients expressing PD-L1 on tumor cells are candidates for anti-PD1 or anti-PD-L1 therapy, immune checkpoint blockade agents that function to restore T cells’ cytotoxic ability in the tumor microenvironment. The approved cut-offs for PD-L1 positivity range from 1 to 50%, depending on the kit used and the cancer indication (2). This wide range and variability of testing make it difficult to draw any reliable conclusion across clinical trials. Another example has been the tissue-agnostic microsatellite instability (MSI) status measured by PCR or IHC staining of key DNA mismatch repair genes. Merck’s anti-PD1 therapy, Keytruda, gained accelerated approval for adult and pediatric cancer patients with advanced solid tumors that are MSI high. This was the first approval based on the genomic alteration of tumors rather than the origin of the primary tumor (3).  A more recent biomarker example is tumor mutational burden (TMB) assessment by next generation sequencing (NGS). Overall TMB is believed to be an important indicator of prognosis and response to immunotherapy. A current challenge in the field is establishing reliable cut-offs for TMB to guide immunotherapy treatment, including the need to determine different cut-off values across tumor types. Friends of Cancer Research has initiated multiple studies in an effort to standardize the TMB measurement across diagnostic panels benchmarked against whole exome as the “ground truth” for accurate somatic mutation measurement, and has outlined key guidelines to ensure consistency across tests (4). These examples, currently in clinical practice, have ushered in a new era of immune checkpoint inhibitors, expanding their efficacy in various cancer types. However, despite their application as clinical biomarkers, PD-L1 status and MSI do not reliably predict individual response, and TMB is merely a surrogate measurement of tumor immunogenicity and neoantigens.

Emerging Genomic Biomarkers of Response

The rapidly changing landscape of predictive biomarkers has been elegantly reviewed by Havel and colleagues (5). It is becoming clear that connecting multiple biomarkers from different cellular pathways and systems will be more advantageous in capturing tumor biology and predicting response. For example, neoantigens trigger the body’s immune system to recognize and destroy tumor cells expressing these foreign peptides that are often absent from normal/non-cancerous cells. These peptides are processed and presented by the HLA complex. Germline HLA zygosity, HLA loss of heterozygosity in the tumor, and somatic alterations impacting HLA genes and molecules of the antigen processing machinery are directly linked to the diversity of neoantigens that can be displayed to the immune system and consequently correlate with response to immunotherapy (6).  Evaluating neoantigen burden alone might not be a strong predictor of response, yet in conjunction with germline and somatic HLA status could yield more predictive power as a biomarker.

Table 1: adapted from Havel JJ et al. Nat Rev Cancer 2019-19-133-50

Biomarkers that provide better information to elucidate the connection between immune system cells and the tumor microenvironment (TME) are also actively being sought. Given the complexities of cancer and the TME in general, it is becoming more appreciated that combination results from a variety of biomarkers is needed to best evaluate patient response to therapy and predict outcome measures. Indeed, a number of recent high impact studies, summarized below, describe an integrated approach to identifying novel biomarkers based on broad assays such as whole exome and transcriptome sequencing.

Roh et al. examined a cohort of patients with melanoma to identify unique molecular and genomic features between responders and non-responders to anti-CTLA4 and anti-PD1 therapy. They explored common melanoma driver genes, HLA alterations, inflammatory signature, mutational load and neoantigen load but observed no significant correlation of these features with response. Looking at additional elements, they identify TCR clonality as a key biomarker of response to anti-PD1 but not anti-CTLA, and focal copy number loss, specifically PTEN loss, as a potential mechanism of resistance to therapy. A more meaningful correlation was observed when combining mutational burden and copy number loss burden as biomarkers (7).

Along a similar vein of comprehensive immunogenomic analysis, Braun et al. examined a cohort of patients with clear renal cell carcinoma treated with anti-PD1 therapy. They evaluated tumor mutational burden, neoantigen load, role of frameshift mutations, zygosity status of Class I HLA genes, and copy number alterations, but did not see any clear associations with improved response. Interestingly, their analysis points to the association of truncating mutations in PBRM1 gene with improved response and survival, and that this genomic alteration was enriched in immune excluded tumors. They also report focal deletions in chromosome 10 that impact PTEN gene which loss correlates with resistance to anti-PD1 therapy (8).

Insight from Immune Parameters in the Tumor Microenvironment

As evidenced by these recent studies, in addition to characterizing genomic alterations, including immune parameters in the classification of cancers to improve predictive value of biomarkers is also key in guiding treatment (9). T cells have been the larger focus of therapies, yet there are now increased efforts to target other cells of the immune system (such as dendritic cells, macrophages, and natural killer cells) for developing clinical therapeutics. Understanding this complexity will require a composite approach. Adding to this complexity is the potential for DNA damage caused by immune cells themselves. Neutrophils, macrophages, and lymphocytes can infiltrate tumors in a manner similar to attacking foreign pathogens. Neutrophils especially release reactive oxygen species and nitric oxide, both potent DNA damaging molecules. Thus, it’s hypothesized that chronic inflammation can increase cancer risk. Analyzing immune cell infiltration and pro-inflammatory mediators can therefore provide additional biomarkers to assess risk. Combination biomarkers have allowed for newer predictive models of tumor burden, such as the recent Immunoscore algorithm which was demonstrated to outperform the classical TNM system (10). This led to the realization that immune infiltration of the TME could be assigned a score, and provide the classification of tumors as “hot” or “cold” based on having a high or low Immunoscore, respectively. Patients with hot tumors are thought to be prime candidates for immune checkpoint therapy. Cold tumors, or those with low immune cell involvement, are very difficult to treat (11). Approaches to better understand the biology of cold tumors in order to convert them into hot and therefore more responsive tumors are best investigated through comprehensive methods such as exome and transcriptome assays.

Personalis has engineered a high-content assay, ImmunoID NeXT Platform®, that combines an analytically validated and augmented exome and transcriptome solution with added unique features to provide accurate and comprehensive immunogenomic characterization of clinical samples. The platform delivers on the biomarkers of today including exome-scale TMB, Microsatellite instability status, and PD-L1 expression. Additionally, this solution delivers an adjusted neoantigen burden based on patient’s specific germline HLA alleles and tumor HLA LOH status, profiling of the antigen processing and presentation machinery, T cell receptor clonality, while also enabling the discovery of novel genomic and immune-related biomarkers representative of the true tumor biology. The ultimate goal is to support the discovery of the next generation of combination biomarkers, improving clinical outcome to various immunotherapies, and designing better strategies for difficult to treat cancers.

Figure 1: Comprehensive Immunogenomics Profiling with ImmunoID NeXT Platform

References

  1. Biomarkers Definitions Working Group. Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clin Pharmacol Ther 2001;69:89-95.
  2. Cottrell, T, Taube, JM. PD-L1 and Emerging Biomarkers in PD-1/PD-L1 Blockade Therapy. Cancer J. 2018 Jan-Feb; 24(1): 41–46.
  3. FDA approves first cancer treatment for any solid tumor with a specific genetic feature [news release]. Silver Spring, MD: US Food and Drug Administration; May 23, 2017. https://www.fda.gov/newsevents/newsroom/pressannouncements/ucm560167.htm. Accessed July 30, 2020.
  4. Merino, D. et al. Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project. J Immunother Cancer. 2020 Mar;8(1):e000147.
  5. Havel JJ, Chowell D, Chan TA. The evolving landscape of biomarkers for checkpoint inhibitor immunotherapy. Nat Rev Cancer 2019;19:133-50.
  6. Chowell, D. et al., Patient HLA class I genotype influences cancer response to checkpoint blockade immunotherapy. Science 10.1126/science.aao4572 (2017).
  7. Roh, W. et al., Integrated molecular analysis of tumor biopsies on sequential CTLA-4 and PD-1 blockade reveals markers of response and resistance. Science Translational Medicine  01 Mar 2017:Vol. 9, Issue 379, eaah3560
  8. Braun, D.A., Hou, Y., Bakouny, Z. et al. Interplay of somatic alterations and immune infiltration modulates response to PD-1 blockade in advanced clear cell renal cell carcinoma. Nat Med 26, 909–918 (2020)
  9. Galon J, Bruni D. Tumor Immunology and Tumor Evolution: Intertwined Histories. Immunity 2020;52:55-81.
  10. Pagès F, Mlecnik B, Marliot F, et al. International validation of the consensus Immunoscore for the classification of colon cancer: a prognostic and accuracy study. Lancet 2018;391:2128-39.
  11. Galon J, Bruni D. Approaches to treat immune hot, altered and cold tumours with combination immunotherapies. Nat Rev Drug Discov 2019;18:197-218.