lec 22 - precision med (molloy)

what is individualized, personalized, precision medicine?

• uses information about genes, proteins and environment to identify patients that most benefit from a particular therapy

• uses biomarkers and companion diagnostics to identify genetic variation among subsets of patients

• uses identification of responders vs non-responders to target treatment

• uses mechanisms to maximize efficacy and safety in pharmacotherapy

goals of precision med

• to correlate specific genetic markers to diseases and therapeutic interventions

• to produce better predictive and diagnostic molecular tests and drugs

• to better select treatments and dosing based on individual need

• right patient, right dose, right time

• increase efficacy and safety

what is a biomarker?

• a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes or pharmacologic responses to a therapeutic intervention

• identification of biomarkers have led to treatment advancements in oncology, HIV/AIDS, neuroscience, CV and others

why are biomarkers important?

• biomarker information is contained in about 60% of FDA approved oncology drug labels between 2015-2019

• across all therapeutic areas, the inclusion of pharmacogenomic biomarkers in drug labels has increased, with about 28% of new drugs in 2020

• biomarker results play an important role in identifying therapeutic outcomes

• biomarkers may categorize responders vs non-responders

• biomarkers maximize safety and efficacy

biomarker categories

• those that assess susceptibility to a disease (prognostic)

  • screening for specific strains of HPV to predict disease susceptibility

  • screening for BRCA gene variants (possible cancer risk)

• those that identify disease and outcome (predictive)

  • predicts the benefit of therapy

  • avoidance of ADRs associated with unnecessary therapy

  • example

    • clopidogrel and 2C19 resistance → some people have genetic variations in CYP2C19 that make clopidogrel less effective

    • warfarin and CYP2C9/VKORC1 → genetic differences in CYP and VKORC affect how people metabolize warfarin and how sensitive they are to it

predictive biomarker components

• efficacy

  • identifying individuals who are more likely to respond to a particular drug

  • efficacy pharmacogenomics

• safety

  • identifying individuals who are less likely to have an adverse effect of a particular drug

  • safety pharmacogenomics

example of predictive biomarker

• patients from at risk populations should be screened for presence of HLA-B*1502 allele

• carbamazepine may bind to HLA-B15 → leading to activation of cytotoxic T cells that attack skin cells and cause SJS/TEN

what are companion diagnostics?

• products and assays used in conjunction with a therapeutic product

• inform treatment selection, initiation, dose customization or avoidance

  • co-development with partner drug

  • post-marketing experience

• FDA guidance

  • use in clinical trials

landscape of oncology personalized medicine

• traditional diagnostic tests in oncology

  • CT

  • MRI

  • PET scans

• new research has broadened use of diagnostics

• evolving discovery of biomarkers leading world of oncology towards personalized medicine

• personalized medicine to have significant impact on development and commercialization of oncology drugs

biomarker methods in cancer

• immunohistochemistry (IHC)

  • application → protein-based assay for detection of expression

  • advantages

    • cheap, rapid, widely available

    • direct visualization of protein expression

  • limitations

    • antibody availability

    • subjective interpretation/quantification

• fluorescence in situ hybridization (FISH)

  • application → hybridization using fluorescent-labeled probes to detect gene copy-number changes or gene rearrangements/fusions

  • advantages

    • relatively simple assay design

    • direct visualization of signals within cells of interest

  • limitations

    • probe availability

    • restricted to specific locus/gene tested

• PCR

  • application → detection of targeted gene mutations, fusions, copy-number alterations, DNA methylation

  • advantages

    • high sensitivity and specificity

    • relative simply assay design

    • relatively low cost

  • limitations

    • limited throughput

    • restricted to targeted genes and regions of interest interrogated

• next-generation sequencing (NGS)

  • application → massively parallel sequencing of multiple genes for detecting mutations, fusions, copy-number alterations

  • advantages

    • high throughput

    • high sensitivity and specificity

    • comprehensive coverage

    • site/tumor-specific applications

  • limitations

    • high complexity

    • bioinformatics requirements

    • longer turnaround time

• gene expression profiling (GEP)

  • application → differential gene expression between tumor/normal or pre/post treated tumor

  • advantages

    • high throughput

  • limitations

    • bioinformatics requirement

    • restricted to targeted genes

schematic for genomics-driven cancer medicine

• pt with tumor is IDed and biopsy is taken

• tumor DNA is extracted and sequenced to look for mutations

• detected mutations are sorted into 3 categories

  • actionable

    • tier 1 → mutations with available FDA approved drugs

    • tier 2 → mutations with drugs currently in clinical trials

  • prognostic

    • mutations that tell us about how aggressive the cancer might be or the patient’s likely outcome

  • variants of uncertain significance

    • mutations we’ve found but do NOT know what they mean clinically

evolution of DNA sequencing tools

• first generation (short read sequencing)

  • sanger sequencing, maxam and gilbert, sanger chain termination

  • infer nucleotide identity using dNTPs, then visualize with electrophoresis

  • 500-1000 bp fragments

• second generation (short read sequencing)

  • 454, solexa, ion torrent, illumnia

  • high throughput from the parallelization of sequencing rxns

  • ~50-500 bp fragments

• third generation (long read sequencing)

  • pacbio, oxford nanopore

  • sequence native DNA in real time with single-molecule resolutions

  • tens of kb fragments on average

types of alterations that can be detected

• point mutation

• indel = small insertions or deletions of DNA bases

• copy number alterations

  • homozygous deletion = both copies of a gene are deleted → complete loss of gene

  • hemizygous deletion = one copy of a gene is deleted → partial loss

  • gain = extra copies of a gene or region → may lead to overexpression

• structural variants

  • translocation breakpoint = parts of 2 different chromosomes break and reattach to each other → can create fusion genes

  • pathogen integrated into genome

potential of tissue based analysis

• genomic and transcriptomic architecture of 2000 breast tumors reveals novel subgroups

innovative medicine: personalized medicine

analyze patients → biomarker diagnostics → results will determine best treatment for each pt

signaling networks regulate the operations of the cancer cell

• growth factor receptors

  • proteins on the surface of the tumor cell

  • EGFR, HER2, c-MET, ALK, FGFR, PDGFR

  • normally, detect signals that tell cell to grow or divide; in cancer, often overactive, sending constant “grow” signals

• non-receptor signaling molecules

  • RAS → RAF → MEK → ERK → cell growth and division

  • PI3K → AKT → mTOR → cell survival, metabolism, growth

  • JAK → STAT → gene transcription

• epigenetic regulators

  • controls how DNA is packaged and which genes are turned on or off

  • cancer cells may alter gene expression WITHOUT changing the DNA itself

• inhibitor of apoptosis proteins

• metabolic modulators

  • cancer cells change their metabolism to support rapid growth:

    • nucleotide synthesis

    • protein & lipid biosynthesis

• angiogenesis

  • cancer cells secrete signals (VEGF, PDGFR) to promote blood vessel growth

• immune checkpoint proteins

  • tumor cells express PD-L1 which binds to PD-1 on T cells which inhibits T cells

  • CTL-A4 inactivates dendritic cells

reliable biomarkers in clinical use - oncology

• HER2 in breast cancer

  • HER2 amplification reliably predicts response to trastuzumab (herceptin) with standardized IHC/FISH testing protocols

• PD-L1 expression

  • used for checkpoint inhibitors but reliability varies by cancer type (e.g. more predictive in NSCLC than others)

• MSI-H/dMMR (high micro-satellite instability/deficient mismatch repair)

  • highly predictive of immunotherapy responses across multiple cancers

• EGFR mutations

  • strong predictor of EGFR inhibitor efficacy in NSCLC

toward precision medicine

• put more science into clinical trials

• use these methods in investigational drug

  • PD measurements

  • molecular diagnostics; candidate approach

  • molecule diagnostics; unbiased approach

cancer is a disease of the genome

• therefore, if we precisely define the cancer genome, we will understand and cure cancer

  • why we must be cautious about such statements

• founder mutations = first genomic mutations

  • these are often lesions that lead to genomic/chromosomal instability and are often NOT fully transforming

• driver mutations = mutations that are required for expression of fully transformed phenotype

  • driver mutations = mutations that we would like to target and inhibit their function

• passenger mutations = these mutations are “collateral damage” resulting from genomic instability and are NOT required for maintaining the transformed phenotype; are “noise” in the system

• since most cancers are rapidly evolving biologic entities, major task to sort out drivers from passengers and these may change over time

therapeutic implications

the things not circles are the hallmarks of cancer and the things in shapes are drugs that fight that hallmark

examples of targeted therapeutics

precision med challenges

precision med pt journey

• neoadjuvant setting (before therapy) → adjuvant therapy (after surgery/therapy) → relapse → metastatic setting

• complementary tools

  • tissue diagnosis → liquid diagnosis

    • tissue = comprehensive

    • liquid = focused specific

oncologic biomarker case study: erbitux (cetuximab) for colorectal cancer

indications and usage

• as a single agent, for treatment of EGFR-expressing metastatic colorectal cancer after failure of both irinotecan- and oxaliplatin-based regiments or in pts who are intolerant to irinotecan-based regimens

• in combination with irinotecan, for treatment of EGFR-expressing metastatic colorectal carcinoma in patients who are refractor to irinotecan-based chemotherapy

EGFR pathway

• transmembrane growth factor receptor belonging to family of HER-related proteins

• ligand binding triggers phosphorylation thru tyrosine kinase activity

• phosphorylation triggers downstream signaling RAS/RAF/MEK/MAPK and PI3K/AKT pathways

  • induction of cancer-cell proliferation

  • blockade of apoptosis

  • activation of invasion and metastasis

  • stimulation of neovascularization

search for a predictive biomarker

• erbitux initially granted approval only for EGFR-expressing patients with mCRC

• 60-80% mCRC tumors express EGFR proteins (as measured by IHC)

• clinical analyses: NO correlation between extent of EGFR expression and response to anti-EGFR mABs

• findings prompted intense research for predictive biomarkers for anti-EGFR mAb

KRAS: the facts

• KRAS = gene that makes KRAS protein which plays a key role in cell signaling

• KRAS mutation occurs in about 40% of CRC patients

• single nucleotide point mutations mainly in codons 12 and 13 of exon 2

• mutation maintained throughout CRC development, progression and metastasis

• KRAS mutation tested by direct sequencing or PCR

early clinical data on KRAS

• 2006 study

  • first reported link between KRAS mutation and lank of response to anti-EGFR mAb

• multiple phase 2 studies of KRAS mutation status

  • cumulative data from 9 studies on 536 patients

  • KRAS mutation detected in 36% of pts

  • response rate (RR), tumor to tumor progression (TTP), progression-free survival (PFS), overall survival (OS) significantly better in WT (wild type) patients

clinical data with cetuximab in KRAS-evaluable patients

• across all 3 studies about 35%-42% of pts had KRAS mutations

• data reinforces the importance of testing for KRAS mutations before using cetuximab → if pt has KRAS mutation, cetuximab is unlikely to help

NCIC Co.18 trial

• in KRAS wild type pts, cetuximab greatly improved OS, PFS, ORR%

• in KRAS mutation pts → NO significant difference whether they got cetuximab or not

KRAS mutation mechanism

• mutated KRAS causes increased signaling despite upstream EGFR inhibition by anti-EGFR drugs hence why drugs like cetuximab do NOT work

tumor with mutant/activated KRAS

• if a tumor has a mutant/activated KRAS, it will grow in the presence of GF inhibitors

conclusions about KRAS

• KRAS = example of how predictive biomarkers can spare patients of using drug WITHOUT potential benefit and with unwanted toxicity and cost

• KRAS example likely to be observed by other therapies

• identification of KRAS as a predictive biomarker = important step toward fulfilling promise of individualized treatment for mCRC

KRAS drug discovery

• after 40 years, “druggable” after all

• 2021 → FDA approval of first KRAS-targeted therapy → sotorasib

• 2022 → adagrasib approved

  • both for mutant lung cancers

conclusions

• ability to obtain full genomic data on a given tumor will allow us to make rational choices for therapy

• functional genomics may provide help in choosing combination therapy

  • combinations will NOT be easy due to enhanced toxicities

• cancer as a chronic disease is NOT a bad thing as long as we recognize rapid development of resistance and clonal evolution

outstanding questions: targeted cancer therapy

• how can we best select ‘biomarker sets’ and properly apply them in clinical treatments of patients to identify optimal target patient subsets, to predict a patient’s response, resistance, and toxicity and to rapidly distinguish between responders and non-responders

• is it possible to screen biomarkers using non-invasive approaches such as circulating tumor cells, circulating DNA, cytokines and chemokines? if not, how can we make technical breakthroughs to fully interpret the information of very limited patients’ biopsies?

• is biomarker-based combinational therapy, that is, a ‘cocktail’ of highly-specific targeted drugs customized to individual patients according to their genetic aberrations, sufficient to largely overcome the resistance of targeted therapy?

  • basically saying: if we design a unique combo of targeted drugs for each pt based on their tumor’s genetic mutations, can we prevent the cancer from becoming resistant and make the treatment work better

• how can innovative biomarker-based clnical design, that is, stratification of patients, assignment of specific drug therapy and adaptive trial designs, increase the translation of targeted drugs from bench to bedside?

• given that the tumor microenvironment has an enormous impact on tumor development, how can we develop models that accurately reflect the tumor microenvironment, in particular the human immune syste, for drug discovery?
  

cancer characteristics

• emerging hallmarks

  • deregulating celular energetics

  • avoiding immune destruction

• enabling characteristics

  • genome instability and mutation

  • tumor-promoting inflammation

immunosuppression and canccer

• immunosuppression is a rate limiting step to effective anti-tumor immunity (for some patients)

• key steps in the cancer-immunity cycle

  • release of cancer cell antigens → tumor cells die and release antigens

  • cancer antigen presentation → dendritic cells or APCs capture these antigens and present them to the immune system

  • priming and activation → APCs activate native T-cells in lymph nodes; anti-CTLA-4 drugs work here by preventing the “off” signals that would normally suppress T cell activation

  • trafficking of T cells to tumors → activated cytotoxic T cells (CTLs) travel thru the blood stream toward the tumor

  • infiltration of T cells into the tumor → cytotoxic T cells move into the tumor tissue

  • recognition of cancer cells by T cells → CTLs recognize cancer cells by binding to antigens presented on their surface; but many cancer cells express PD-L1 which binds to PD-1 on T cells, turning them off; anti-PD1/PD-L1 drugs block this signal off so T cells can stay active

  • killing of cancer cells → CTLs kill the cancer cells

• where immunosuppression happens

  • tumor-induced immunosuppression can interfere at multiple steps:

    • prevent antigen presentation

    • stop T-cell activation

    • prevent T cells from reaching tumors

    • supress T cell killing ability

rise of immunotherapy

• new therapeutics

  • CART cell therapy

  • customized vaccines

• process of immunotherapy

  • T cells are collected from the patient’s blood

  • T cells are genetically reprogrammed to find and attack cancer cells; T cells are grown in large numbers in lab

  • reprogrammed T cells are infused into the patient

immune checkpoint inhibitors

• CTLA-4 blockage (e.g. ipillimumab)

  • CTLA-4 checkpoint protein prevents dendritic cells from priming T cells to recognize tumors → inhibitor drug blocks the checkpoint

• PD-1 blockage (e.g. nivolumab, pembrolizumab, atezolizumab)

  • PD-1 checkpoint protein prevents T cells from attacking cancer cells → inhibitor drug allows T cells to act

broad activity for anti-PD-L1/PD-1 in human cancer

• basically affects a bunch of cancer like

  • head and neck cancer

  • lung cancer

  • gastric

• broad activity but only subset of patients benefit → ~10-30%

combo of immunotherapeutics or immunotherapeutics with SOC/targeted therapies

• shows that best survival outcomes = immunotherapy + targeted/chemotherapy

• agents must be safe in combination with anti-PD-L1 (immunotherapy)

• targeted/chemotherapy should NOT interfere with immune response or immunotherapeutic MOA

conclusions

• personalized medicine can be regraded as the 21st century’s answer to the rational use of drugs → right drug for the right pt

• key drivers in personalized medicine:

  • molecular diagnostics

  • academic groups

  • patient advocacy groups

  • authorities and health insurance companies

• implementing personalized medicine into clinical practice will be a challenge that needs to address various other issues such as regulatory requirements, reimbursement, education and logistics