class: center, middle, inverse, title-slide # NGS interpretation and reporting: in the view of pathologists ### Jun Kang ### Catholic University of Korea ### 2022 02 12 --- # Contents * Allele frequency * Variant pathogenecity ??? * 병리의사는 발견된 변이에 대한 평가를 하고 보고서에 변이에 대한 임상적 의미를 기록하게 됩니다. * 변이에 대한 종합적 판단을 위해서는 Bioinformatics에 대한 이해가 필요한 부분이 있습니다. * NGS 결과를 받아보면 variant 정보가 많이 있습니다. * 그리고 임상적 상황과 맞추다 보면 발견된 variant에 대한 질문도 많이 있습니다 * 질문에 답할 때 allele frequency 가 참고 되는 경우가 많이 있습니다. * 예를 들면 + Somatic vs germline + False positive + Allele specific copy number (LOH) * 이번 시간은 첫번째로 Allele frequency가 어떻게 나오게 되는지 이해하고자 합니다. * 두 번째로는 tumor suppressor gene의 변이는 유전자 기능이상을 발생하는지 확인이 필요합니다. * 유전자 기능 변화가 변이 별로 다르기 때문에 어떠한 기준으로 변이의 병원성 여부를 확인하는 지 알아보고자 합니다. --- ### Questions * Somatic vs germline * False positive or false negative (cut-off) * Driver vs passenger (ultra-hypermutation, POLE) * Tier ??? * 발견된 변이들을 보게 되면 다음과 같은 질문들을 하게 됩니다. * 이 경우 allele freqency를 먼저 확인해 보면 판단에 도움이 됩니다. --- ## Allele frequency * BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T) * Variant allele frequency 63.84% * Tumor cell percentage: 70% * BRCA negative in blood sample ??? 예를 보겠습니다. 난소암 환자에서 발견된 BRCA1 변이이고 stop codon을 만드는 nonsense variant입니다. nonsense variant이기 때문에 pathogenic varaint로 생각됩니다. 변이 allele frequency는 63%이고 병리의사가 측정한 종양세포 비율은 70%입니다. 환자는 혈액 BRCA 검사에서 음성이기 때문에 종양검체에서 시행한 NGS에서 발견된 이 변이는 germline이 아니라 somatic으로 생각됩니다. 실제로 임상 선생님으로부터 이 변이에 대한 문의를 받았습니다. --- ### Questions > Is it possible **64% allele frequency** of somatic mutation? ??? 질문은 allele frequency가 63%로 50%에 가까운데 somatic varaint일 수 있느냐라는 것입니다. --- ### Questions > Can we determine whether the variant is germline or somatic using allele frequency and tumor percentage in tumor only test? ??? 질문을 바꿔보면 allele freqeuncy로 tumor only NGS 검사에서 germline과 somatic을 구분할 수 있느냐 입니다. --- class: center # Variant allele frequency in clinical tumor sample `$$Allele \space frequency \approx Read \space count \space proportion$$` ??? variant allele frequency는 무엇일까요? allele의 read count의 비율입니다. 여러 예를 통해서 allele frequency의 의미에 대해서 알아보고자 합니다. --- class: center, middle  ??? 동그란건 종양세포 타원형은 정상세포입니다. 종양세포의 비율은 25% 1/4입니다. 각각 두개의 allele이 있고 종양세포에만 변이가 있는 somatic 변이인 경우 variant allele frequency는 어떻게 될까요? 전체 allele은 8개이고 그 중 하나만 변이이니까 15.5% 1/8 이 됩니다. --- class: center ### Somatic variant, Two copy, Tumor cellularity 50%  Variant allele frequency? ??? 다음은 이전과 같지만 종양세포 비율이 50%로 두배가 되었습니다. 이 varaint allele frequency는 네개중 하나 25%입니다. 두 예에서 variant allele frequency는 종양세포 비율의 절반으로 생각할 수 있습니다. 종양세포의 두 allele중 하나만 변이가 생겼기 때문에 절반으로 볼 수 있습니다. --- class: center ### Germline variant, Heterozigosity, Two copy, Tumor cellularity 25%  Variant allele frequency? ??? 이 경우는 종양세포 비율이 25%인데 germline variant 인 경우 모든 세포의 두 allele중 절반에서 변이가 있기 때문에 50%가 됩니다. --- class: center ### Germline variant, Heterozigosity, Two copy, Tumor cellularity 50%  Variant allele frequency? ??? 종양세포 비율이 커져도 50%가 됩니다. --- # Proposition Germline, homo | Germline, hetero | Somatic :-----: | :----: | :----------------------: 100% | 50% | half of tumor cellularity, `\(\le 50\%\)` ??? 흔히 germline variant allele frequency는 50% somatic variant allele frequency는 종양세포 비율의 절반이라고 말 하기도 합니다. 하지만 이런 경우는 추가 조건이 필요합니다. --- class: center ### Germline variant, Heterozigosity, One copy, LOH, Tumor cellularity 50%  Variant allele frequency? ??? 종양세포는 allele이 두개가 아닌 allele copy number를 가질 수 있습니다. 이 경우 loss of heterozygosity로 정상 allele이 deletion 되고 variant copy만 가지고 있는 tumor suppressor gene의 종양발생 기전으로 알려진 two hit theory에 해당하는 경우입니다. germline 변이가 있는 경우 LOH가 동반되어 종양이 생기게 됩니다. 이 경우 germline 변이이지만 allele frequency는 50%가 아니라 66%이 됩니다. 종양세포에서 정상 allele이 없어졌기 때문입니다. --- class: center ### Germline variant, Heterozigosity, One copy, LOH, Tumor cellularity 25%  Variant allele frequency? ??? 이 경우는 상황은 같지만 종양세포 비율이 25%로 줄어 variant allele frequency는 약간 줄어 57%이 되었습니다. --- class: center ### Somatic variant, One copy, LOH, Tumor cellularity 50%  Variant allele frequency? ??? somatic variant 에 LOH가 동반되면 어떻게 될까요? 이 경우 9개중 3개로 33%인데 종양비율의 절반인 25% 보다 높습니다. --- class: center ### Somatic variant, Amplification, Tumor cellularity 25%  Variant allele frequency? ??? LOH가 아니라 amplification이 동반되면 더 많이 올라가게 됩니다. --- # Variant allele frequency in clinical tumor sample * Germline vs somatic * Tumor cell proportion * Loss of heterozygosity * Copy number ??? variant allele frequency는 gerline인지 somatic인지 ....에 따라 결정되게 됩니다. --- # Allele frequency in Somatic vs Germline in tumor only sample * BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T) * Variant allele frequency 63.84% * Tumor cell percentage: 70% ??? 처음 실제 발견된 변이를 다시 살펴보면 BRCA1 유전자는 tumor suppressor gene이며 난소암에서 LOH가 있을 것으로 생각됩니다. 그러면 variant allele frequency가 50%가 넘는 것을 설명할 수 있습니다. 하지만 이게 germline인지 somatic인지 알 수 있을까요? --- # Allele frequency fomulas [1] `\(AF_{germline}\)` `\(=\)` `\({pV+1-p} \over {pC+2(1-p)}\)` `\(AF_{somatic}\)` `\(=\)` `\(pV \over pC+2(1-p)\)` * Given copy number (C) * Variant allele count (V) * Sample purity (p) * Variant status (somatic or germline) ??? allele frequency를 공식으로 만들어 보면 이렇게 됩니다. --- # Limited information What we know | What we don't know :-------------:|:-----------------: Tumor cell percentage | Copy number Variant allele frequency | LOH ??? 종양세포비율과 variant allele frequency는 알 고 있고 copy number와 LOH는 모른다고 가정하면 --- # Allele frequency * BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T) * Variant allele frequency 63.84% * Tumor cell percentage: 70% -> 80% * Somatic (supposed) * LOH (supposed) * One copy (supposed) ??? 이 경우 다음과 같이 가정하고 variant allele frequency를 계산해보겠습니다. --- class: center ## Allele frequency  `\(AF_{somatic}\)` `\(=\)` `\(pV \over pC+2(1-p)\)` `\(= 0.67\)` ??? 계산된 allele frequency와 실제 NGS 값과 비슷합니다. --- # Allele frequency * BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T) * Variant allele frequency 63.84% * Tumor cell percentage: 70% -> 60% * Germline (supposed) * LOH (supposed) * One copy (supposed) ??? 이번에는 종양세포 비율을 60%로 낮추고 germline으로 가정하고 계산을 해보겠습니다. --- class: center ## Allele frequency  `\(AF_{germline}\)` `\(=\)` `\({pV+1-p} \over {pC+2(1-p)}\)` `\(= 0.71\)` ??? 71%로 실제값 63%와 차이가 있지만 비슷한 값이 나왔습니다. 실제로는 정확한 종양세포 비율과 allele specific copy number를 알 지 못하기 때문에 단순히 variant alllele frequency만으로 germline과 somatic을 구분하기는 어렵습니다. --- # Caveat * Strand bias * Heterogeneity ??? strand bias 나 heterogeneity가 있는 경우에는 allele frequency해석에 주의가 필요합니다. --- # Allele specific copy number  ??? 반대로 allele frequency와 DNA양인 coverage를 알고 있다면 allle specific copy number를 추정할 수 있습니다. --- # Variant pathogenecity ??? 이제부터는 variant pathogenecity를 어떻게 판단하는지 살펴보겠습니다. --- # Question * 76/M, Lung adenocarcinoma * NM_000059.3(BRCA2):c.5683G>A (p.Glu1895Lys) ??? 76세 폐암 환자에서 BRCA2 missense variant가 발견되었되었습니다. --- class: my-one-page-font >This sequence change replaces glutamic acid with lysine at codon 1895 of the BRCA2 protein (p.Glu1895Lys). The glutamic acid residue is weakly conserved and there is a small physicochemical difference between glutamic acid and lysine. This variant is present in population databases (rs146351301, ExAC 0.009%). This variant has been reported in individuals affected with breast cancer (PMID: 25682074, 20104584, 27257965, 28664449). In the literature, this variant is also known as 5911G>A. ClinVar contains an entry for this variant (Variation ID: 142307). Algorithms developed to predict the effect of missense changes on protein structure and function output the following: SIFT: "Tolerated"; PolyPhen-2: "Benign"; Align-GVGD: "Class C0". The lysine amino acid residue is found in multiple mammalian species, suggesting that this missense change does not adversely affect protein function. These predictions have not been confirmed by published functional studies and their clinical significance is uncertain. In summary, the available evidence is currently insufficient to determine the role of this variant in disease. Therefore, it has been classified as a Variant of Uncertain Significance. <footer>Sherloc Invitae</footer> ??? Clinvar에서 이 변이 평가 기록중 invitae에서 작성한 걸 보보겠습니다. BRCA2단백질 코돈 1895의 glutamic acid가 lysine으로 바뀌었고 glutamic acid는 weakly conserve 되었고 population에서 0.009%에서 발견됩니다. 유방암환자에서 발견 된적이 있고 in-silico program prediction은 benign으로 예측 되었습니다. functional study는 시행되지 않았습니다. 이런 증거들을 바탕으로 invitae는 variant of uncertain significance로 평가 하였습니다. ---
??? 이 자료는 vcf에서 확인 할 수 있고 NGS 결과 확인 프로그램에서도 기능이 제공되기도 해서 쉽게 확인 할 수 도 있습니다. 자료를 살펴보면 Invitae에서 평가하였던 근거 자료들 임을 알 수 있습니다. --- ## Standards and Guidelines for the Interpretation of Sequence Variants * To describe variants identified in Mendelian disorders * **American College of Medical Genetics and Genomics (ACMG)** * ENIGMA BRCA1/2 Gene Variant Classification Criteria * International Agency for Research on Cancer (IARC) ??? Mendelian disorder를 일으키는 변이인지 확인하는 목적이고 pathogenic variant를 가지면 유전성 질환으로 진단을 할 수 있습니다. 많은 hereditary cancer syndrome에 여러 tumor suppressor gene이 알려져 있고 이 경우 다양한 변이들이 유전질환의 원인 변이가 되고 이를 평가하는 것에 어려움이 많이 있습니다. 일관된 평가를 위해 pathogenesity를 평가하는 여러기준이 있지만 ACMG guideline을 소개하려고 합니다. --- # Categories of interpretation of variants * Pathogenic * Likely-pathogenic * Uncertain (VUS) * Likely-benign * Benign ??? 다음과 같은 5가지 카테고리가 있습니다. --- # Let's guess the evidences ??? 어떤 정보를 확인하면 변이가 단백질 기능 상실을 일으키는 지 알 수 있을까요? --- # Famly pedigree  ??? 우선 가계도를 생각해볼 수 있습니다. 가계 내 질병이 생긴 환자에서 해당변이가 더 자주 발견되면 pathogenic variant로 생각할 수 있습니다. --- # Segregation data (BS1, PP1) * Caveat + linkage disequilibrium + Penetration rate + Difficult statistical evaluation ??? 하지만 다른 위치에 실제 pathogenic variant가 있고 발견된 변이가 같이 유전되는 linkage disequilibrium 이나 유전력이 낮은 경우 판단이 어렵습니다. --- # Population data  ??? 다른 증거로는 population 정보가 있습니다. 대규모 집단 유전체 자료가 공개되어 있고 여기에서 common snp이라면 benign variant로 판단할 수 있습니다. --- # Population data * 5%: benign stand alone (BA1) * 0.5-5% (BS1) * The first time observed variant! (Absent in population DB, PM2) ??? 대규모 집단 유전체 자료에서 빈도가 5%이상인 common snp인 경우 benign variant로 판단할 수 있습니다. --- # Null variant * Frameshift, Nonsense, canonical +-1 or 2 splicing site, initiation codon * Caveat: LOF variants at the extreme 3′ end of a gene * Caveat: presence of multiple transcripts ??? 유전자의 기능 소실이 예상되는 Frameshift, Nonsense 등 null variant는 강력한 pathogenic variant의 증거가 됩니다. --- # Computational (in silico) data * PolyPhen2, SIFT, MutationTaster, etc * Mutational hot spot and/or critical and wellestablished (PM1) * Protein length changes due to in-frame deletions/insertions and stop losses functional domain (PM4 BP3) * Novel missense at the same position (PM5) ??? 변이의 기능 소실을 예측해주는 여러 프로그램 예측결과도 어느정도 증거로 생각할 수 있습니다. --- # Other evidence * de novo variants (PS2 PM6) * Functional studies (PS3 BS3) * Allelic data (BP2 PM3) ??? 그 외 다른 증거들로 ... --- # Evidences of interpretation * Population data * Computational data * Functional data * Segregation data * De novo data * Allele data * Other databases * Other data ??? ACMG guideline에서는 8가지 evidence category를 바탕으로 판단하게 됩니다. --- ##### 27 variant attributes  ??? 27 variant attributes가 있지만 실제로 판단이 가능한 항목들은 제한적입니다. --- class: small-font # Characteristics of BRCA1/2 * LOF known mechanism of disease (for PVS1) * Mode of inheritance (for PM3/BP2) * AD/AR (BRCA2) * Missense pathogenic (for PP2/BP1) * BRCA2 1% * Hot spot or critical/well-established functional domain (for PM1) * BRCA2, Helical (2479-2667), OB (2670-2799 and 3052-3190), Tower (2831-2872) ??? 실제 예를 살펴보겠습니다. BRCA1/2 유전자의 특징을 확인하여 해당하는 항목들을 확인할 수 있습니다. ... --- class: small-font ## NM_000059.3(BRCA2):c.5683G>A (p.Glu1895Lys) * LOF known mechanism of disease (for PVS1) * Mode of inheritance (for PM3/BP2) * AD/AR (BRCA2) * Missense pathogenic (for PP2/**BP1**) * **BRCA2 1%** * Hot spot or critical/well-established functional domain (for PM1) * BRCA2, Helical (2479-2667), OB (2670-2799 and 3052-3190), Tower (2831-2872) ??? 이 중 missense variant인 점은 benign supporging evidence입니다. BP2가 benign supporting evidence의 약자입니다. 나머지 항목은 해당사항이 없습니다. --- ### Characteristics of varant (NM_000059.3(BRCA2):c.5683G>A (p.Glu1895Lys)) * ClinVar (Uncertain significance (Last evaluated: Nov 1, 2015)) (PP5, BP6) * Population AF: `\(8.29e^{-06}\)` (PM2, BA1, BS1) * Insilico SIFT 1.0, phyloP -0.72, PolyPhen-2 0.004 **BP4** ??? 다음 확인 항목들 중에서 insilico data도 benign supporting evidence입니다. --- # Functional assay (BS3) * Findlay (2018) Nature 562: 217 PubMed: 30209399 * Guidugli (2013) Cancer Res 73: 265 PubMed: 23108138 * Biswas (2011) Blood 118: 2430 PubMed: 21719596 * Becker (2012) Breast Cancer Res Treat 135: 167 PubMed: 22729890 ??? Functional assay에서는 해당변이 자료를 찾을 수 없었습니다. --- class: my-one-page-font >This sequence change replaces glutamic acid with lysine at codon 1895 of the BRCA2 protein (p.Glu1895Lys). The glutamic acid residue is weakly conserved and there is a small physicochemical difference between glutamic acid and lysine. This variant is present in population databases (rs146351301, ExAC 0.009%). This variant has been reported in individuals affected with breast cancer (PMID: 25682074, 20104584, 27257965, 28664449). In the literature, this variant is also known as 5911G>A. ClinVar contains an entry for this variant (Variation ID: 142307). Algorithms developed to predict the effect of missense changes on protein structure and function output the following: SIFT: "Tolerated"; PolyPhen-2: "Benign"; Align-GVGD: "Class C0". The lysine amino acid residue is found in multiple mammalian species, suggesting that this missense change does not adversely affect protein function. These predictions have not been confirmed by published functional studies and their clinical significance is uncertain. In summary, the available evidence is currently insufficient to determine the role of this variant in disease. Therefore, it has been classified as a Variant of Uncertain Significance. <footer>Sherloc Invitae</footer> ??? 이전에 보았던 invitae에서 시행한 variant interpretation을 보면 방금 했던 과정들은 설명하고 있습니다. --- # BRCA2 c.10150C>T p.Arf3384Ter ??? 이번에는 다른 변이를 생각해 보겠습니다. 이 변이는 BRCA2 단백질 2884 코돈에 stop codon이 생겨 nonsense variant 입니다. --- ## KOBRA [2]   ??? 한국인 유전성유방암 연구에서 pathogenic으로 발표되었습니다. --- ## End truncation  ??? 하지만 BRCA2 유전자를 보면 전체 아미노산 길이가 3418로 절단은 단백질 말단 3384 코돈에서 발생해서 실제 유전자 기능이상을 일으키지 않을 가능성도 있습니다. --- # Clinvar BRCA2 c.9976A>T (p.Lys3326*) variant, located upstream of this variant and also in the last exon of the gene, is a known benign variant. ??? clinvar에서는 nonsense variant이지만 benign으로 분류하고 있습니다. --- # Conclusions * Allele frequency * ACMG guideline ??? 지금까지 allele frequency와 ACMG guideline에 대해 말씀 드렸습니다. 감사합니다. --- class: my-one-page-font # References [1] J. X. Sun, Y. He, E. Sanford, et al. "A computational approach to distinguish somatic vs. germline origin of genomic alterations from deep sequencing of cancer specimens without a matched normal". En. In: _PLOS Computational Biology_ 14.2 (2018), p. e1005965. ISSN: 1553-7358. [2] E. Kang, M. Seong, S. K. Park, et al. "The prevalence and spectrum of BRCA1 and BRCA2 mutations in Korean population: recent update of the Korean Hereditary Breast Cancer (KOHBRA) study". En. In: _Breast Cancer Res Treat_ 151.1 (5. 2015), pp. 157-168. ISSN: 1573-7217. <style type="text/css"> .remark-slide-content { font-size: 24px; padding: 1em 4em 1em 4em; } .my-one-page-font { font-size: 18px; } .beginners{ visibility: hidden; } .intermediate{ visibility: hidden; } .advanced{ visibility: visible; } { max-width: 100%; } </style>