Biomarkers for Homologous Recombination Deficiency in Cancer

 Note for: Biomarkers for Homologous Recombination Deficiency in Cancer

Doi: 10.1093/jnci/djy085

Background:

-          Defective in DNA repair is hallmark of cancer (Problem)

-          HR-defective is of interest in clinical level due to its sensitivity to PARPi

-          HR-defective measurement is important for the clinical trial design

-          Identified method for HRD -- > varied and controversial

-          If we understand more in this measurement, it would help the clinical trial design

Aim:

-          Summarize biology and clinical validation of current methods to measure HRD

-          Helping in decision making to treat patients

-          Informing indicators for HRD

o   Germline BRCA1/2 mutation status

o   Clinical response to Pt-based therapy

o   Newer assay undergoing clinical validation

§  Somatic mutation in HR genes

§  Genomic scar assays using array-based comparative genomic hybridization (aCGH)

§  Info derived from NGS

·         Single nucleotide polymorphism analysis

·         Mutational signatures

§  Transcriptional profiles of HRD

§  Phenotypic functional assays of protein expression and localization

-          Highlight

o   Strength and weakness of each assays

-          HRD biomarker

o   Use with either BRCA1/2 defective or BRCA1/2 wild type

Main text

Approved HRD biomarkers for PARP inhibitor use

-          Germline BRCA (gBRCA) mutation is the best clinical biomarkers

-          Myriad Genetics BRACAnalysis CDx platform -- > FDA-approved to detect gBRCA mutation

-          No current approved diagnostic assay for HRD based on germline mutation of other homologous recombination genes

-          Variants of uncertain significance (VUS) in other genes -- > unclear

-          Biomarkers should not be focused on only BRCA wild-type but BRCA mutant population -- > there might be phenotypic variants and reversions

Platinum sensitivity as surrogate biomarker for HRD

-          Pt-sensitivity -- > feature for HRD

-          Pt-sensitivity -- > used as surrogate clinical index for prediction of efficacy to PARP inhibition

-          Some Pt-resistant patients has shown PARPi sensitivity

Novel biomarkers of HRD

-          Somatic mutation in HR genes

o   In vitro criteria – predicting sensitivity to PARPi

§  Defects in recombination substrate assays

§  Inability to form RAD51 foci formation upon exposure to DNA damage

§  In vitro sensitivity to platinum salts and PARPi

o   Cancer-associated mutations in genes related HR -- > all of these below genes face uncertainty over functional significane

§  PALB2

§  BARD1

§  BRIP1

§  RAD51B

§  RAD51C

§  RAD51D

§  ATM

§  FAAP20

§  CHEK2

§  FAN1

§  FANCE

§  FANCM

§  POLQ

-          HR-related genes are synthetic lethal with PARPi

o   Different in tumor type (context-specific clinical evaluation)

-          There are the attempt to do multiomics (genomic, transcriptomic, proteomic) as well as functional assay to clarify HRD tumor

-          Genomic scar assays

o   Cancer genome often harbor chromosomal aberrations -- > result from defective DNA repair -- > increased gross chromosomal rearrangements

o   Thus, leading to evaluated the “genomic scar” through hig-throughput profiling technique

§  Array-based comparative genomic hybridization (aCGH)

§  SNP genotyping

§  NGS

-          Array-based comparative genomic hybridization (aCGH)

o   Detecting genomic copy number variation in tumors

o   Can be used with FFPE samples

o   aCGH assays have not been evaluated in the context of PARP inhibition

-          SNP-based genomic scar assays

o   Three SNP-based assays developing to quantify the extent of chromosomal abnormalities

§  Telomeric allelic imbalance (TAI)

§  Loss of heterozygosity (LOH)

§  Large-scale transition (LST)

o   Developing by using training cohort of tumors from BRCA1/2 mutated patients

§  In silico analysis; 5371 tumors from 15 cancer types (in TCGA), Pt is first line regiment

-          

 Mutational signatures

o   Signature 3 -- > can identify HRD

o   HRDetect -- > identify BRCA-deficient tumors at high sensitivity and specificity from FFPE

o   HRDetect algorithm -- > has to be validated in clinical trial of PARP inhibition

-          Limitation of genomic scar assays

o   Emergence of resistance pathways would not remove the existing “scar” of prior HRD

o   Reversion or secondary mutations have been described to restore HR

o   BRCA mutation leaves scar -- > upon reversion -- > regain HR

o   Assays that report a “real-time” index of HR in tumor sample would be of clinical interest and value

-          Real-time indicators of HRD

o   Genomic alteration -- > reflect past events

o   RNA and proteins vary dynamically in

§  Quantity

§  Localization

·         These two would reflect the real-time HRD indicator

o   Transcriptional profiles

§  Gene expression profiling -- > oncotypeDx

§  60 gene signature clusters into BRCA-like and non-BRCA-like and BRACAness profiles

§  Most assays come from collecting the data from the cancer cohort

o   Protein expression

§  Depletion of ATM -- > associated with PARPi sensitivity in vitro

§  Multiplex analyses of DNA repair proteins in clinical trial -- > herald new information on biomarkers of HRD

o   Functional assays

§  Measuring all proteins of interest within a pathway for each cancer sample

§  Propose to measure a single downstream event -- > can reflect proficiency of multiple upstream components of homologous recombination

·         Quantification of RAD51 foci formation -- > inability of cells to form RAD51 foci -- > common feature of HRD

o   RAD51 foci in S/G2 formation has been evaluated in FFPE breast cancer tissue at 24 hrs post chemotherapy using geminin a marker for S and G2 cells

o   RAD51-low score -- > more common in TNBC than in other subtypes

o   Major challenges -- > reliable quantitation of “foci” in FFPE samples -- > fixation artifacts -- > common

o   Lacking of availability of post-treatment biopsies in routine clinical samples -- > baseline level of RAD51 (and other DDR) foci -- > not relevant to relative increase in their numbers upon DNA damage

o   Tissue context, modality of DNA damage used, timing of RAD51 foci assessment, resolution of microscopy system -- > causing large differences of foci assessment between studies

o   False positive of RAD51 foci formation, but still HR defect

§  occur due to downstream of ssDNA-RAD51 filament has defect -- > RAD51AP1, Pol-eta

§  loss of upstream DDR components of MRN complex and ATM

·         RAD51 foci formation --- > biologically useful marker -- > not easily quantifiable in clinical material

o   Propose measurable marker -- > replication fork stabilization -- > promising phenotypic marker of PARPi sensitivity

Conclusion

-          Generic touchstone biomarker for HRD -- > key to expanding therapeutic utility of HRD-targeting agents across a broad spectrum of tumor types

-          Complexity of HR – thus one marker is not enough to identify defect in HR -- > define HRDness

o   Pt-sensitivity

o   HRDetect positivity (mRNA/protein expression profile)

o   Decreased replication fork stability

-          Clinical application

o   Rapid

o   Precise readouts

-          Nature of genomic instability in HRD tumor cells

o   Increase the rate of therapeutic evasion (resist to treatment) through the clonal evolution

§  Thus, it is better to treat as early as possible

o   In the late stage, it is hard to validate the biomarker for HRD – due to tumor heterogeneity and clonal evolution

§  Thus, re-evaluation of tumor HR status in patients is important

§  HRD biomarkers based on tissue biopsy

·         Serial and multiple spatially distinct samples

§  HRD biomarker based on liquid biopsy

§  Help to identify right patients at  right time for right therapeutic -- > for HRD-directed precision medicine