PO.BCS01.11 · 生物信息与计算
PlasmaCHORD——一种用于在液体活检中识别克隆性造血变异的机器学习方法
PlasmaCHORD- A machine learning method for identifying clonal hematopoiesis variants in liquid biopsies
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
引言：通过液体活检对循环肿瘤DNA（ctDNA）进行基因组分析已成为临床肿瘤学中的重要诊断方法。然而，与克隆性造血（CH）相关的变异的检出是一个主要的混杂因素，会削弱液体活检的临床应用价值。降低血浆NGS中来自CH的生物学噪声的策略包括对配对的白细胞（WBC）DNA进行深度测序和/或对肿瘤组织进行测序。虽然这些方法能有效区分大多数CH变异，但对额外生物标本和测序的需求增加了成本并限制了可行性。
方法：我们利用一个训练队列（来自225例I-IV期实体瘤患者的ctDNA NGS中鉴定出的426个变异），开发了plasmaCHORD，这是一个机器学习模型（MLM），纳入了片段组学（fragmentomic）、变异层面和患者层面的特征，以区分通过固定基因panel杂交捕获NGS检测到的突变是肿瘤来源还是CH来源。模型性能通过与根据配对WBC和肿瘤NGS确定的每个血浆变异的参考来源进行比较来评估。在锁定模型参数后，我们将plasmaCHORD应用于一个独立验证队列（在114例转移性癌症患者中检测到的1412个血浆变异），以及应用于一项前瞻性液体活检指导的临床试验（NCT05585684）入组患者的cfDNA NGS。
结果：plasmaCHORD在训练集中以高准确度预测肿瘤来源与CH来源（交叉验证AUC=0.94），优于变异等位基因频率和经典CH基因等单一特征。当限定于由3-5个突变reads支持的突变DNA片段时，模型性能仍然稳健（AUC=0.84）。plasmaCHORD以0.5的分值作为区分肿瘤来源与CH来源变异的临界值进行锁定评估。在独立验证队列中，锁定的模型保持了相似的总体准确度（AUC=0.9），敏感性为82%，特异性为80.3%，准确度为80.2%。我们的方法在对通常不与CH相关的临床可操作基因（包括AKT1、ATM、BRCA1、BRCA2和EGFR）进行变异来源分类方面表现出高度可靠性，同时也能对在实体和血液系统恶性肿瘤中均可遇到的TP53突变进行细胞来源判定。该性能在不同癌种、测序平台、突变类别以及广泛的等位基因分数范围内均保持一致。当应用于精准肿瘤学临床试验背景下具有临床挑战性的病例时，plasmaCHORD精确地确定了变异来源，避免了与基因型靶向治疗之间的错配。
结论：plasmaCHORD是一个多特征机器学习分类器，能够显著提升在常规仅血浆NGS中识别真正肿瘤变异的能力，通过最大限度减少由CH引起的误读，满足了实施液体活检指导治疗中的一项关键需求。
查看英文原文 English abstract
Introduction: Genomic profiling of circulating tumor DNA (ctDNA) through liquid biopsies has become an important diagnostic method in clinical oncology. However, detection of variants related to clonal hematopoiesis (CH) is a major confounder that impairs the clinical utility of liquid biopsies. Strategies that reduce biological noise from CH in plasma NGS include deep sequencing of matched WBC DNA and/or tumor tissue sequencing. While these methods effectively distinguish most CH variants, the need for extra biospecimens and sequencing raises costs and limits feasibility.
Methods: Using a training cohort of 426 variants identified in ctDNA NGS from 225 patients with stage I-IV solid tumors, we developed plasmaCHORD, a machine learning model (MLM) that includes fragmentomic, variant, and patient-level features to distinguish between tumor- and CH-origin for mutations detected by fixed gene panel hybrid capture NGS. Model performance was assessed by comparison to the reference origin of each plasma variant determined from matched WBC and tumor NGS. Following locking the model parameters, we applied plasmaCHORD to an independent validation cohort of 1,412 plasma variants detected in 114 patients with metastatic cancers, as well as to cfDNA NGS from patients enrolled in a prospective liquid biopsy-informed clinical trial (NCT05585684).
Results: PlasmaCHORD predicted tumor versus CH-origin in the training set with high accuracy (cross-validated AUC=0.94), outperforming individual features such as variant allele frequency and canonical CH genes. Model performance remained robust when restricted to mutant DNA fragments supported by 3-5 mutant reads (AUC = 0.84). plasmaCHORD was locked for evaluation using a score of 0.5 as cutoff for distinguishing tumor- versus CH-origin variants. In the independent validation cohort, the locked model maintained similar overall accuracy (AUC=0.9) with a sensitivity of 82%, specificity 80.3% and accuracy of 80.2%. Our approach was shown to be highly reliable in classifying variant origin in clinically actionable genes not canonically associated with CH, including AKT1, ATM, BRCA1, BRCA2, and EGFR, as well as adjudicating cellular origin for TP53 mutations that are encountered in both solid and hematologic malignancies. Performance was consistent across cancer types, sequencing platforms, mutation classes, and a wide range of allele fractions. When applied to clinically challenging cases in the context of a precision oncology clinical trial, plasmaCHORD precisely determined variant origin, preventing mismatches with genotype-targeted therapies.
Conclusions: plasmaCHORD, a multi-feature machine-learning classifier, can significantly enhance the ability to identify bona fide tumor variants in routine plasma-only NGS, addressing a critical need in implementing liquid biopsy-guided therapy by minimizing misinterpretation caused by CH.
利益披露 Disclosure
D. J. Rabizadeh, None.
J. V. Canzoniero,
Foundation Medicine ).
AstraZeneca ).
I. Ziakas, None..
J. Wehr, None..
A. Balan, None.
S. C. Scott,
AstraZeneca Independent Contractor, ).
Black Diamond Therapeutics ).
Janssen Independent Contractor, ).
Daiichi Sankyo Independent Contractor.
EMD Serono Independent Contractor.
Foundation Medicine Independent Contractor.
Genentech Independent Contractor.
Merus Independent Contractor.
Regeneron Independent Contractor.
Tempus Independent Contractor.
G. Pereira, None.
V. K. Lam,
Pfizer Independent Contractor.
Genentech/Roche Independent Contractor.
Iovance Biotherapeutics Independent Contractor.
Anheart Therapeutics Independent Contractor.
Takeda Independent Contractor.
Seattle Genetics Independent Contractor, ).
Bristol Myers Squibb Independent Contractor, ).
AstraZeneca Independent Contractor, ).
Guardant Health Independent Contractor.
GlaxoSmithKline ).
Merck ).
C. M. Lovly,
AbbVie Independent Contractor.
Amgen Independent Contractor.
AnHeart Independent Contractor.
Astra Zeneca Independent Contractor.
Black Diamond Independent Contractor.
BMS Independent Contractor.
Jazz Independent Contractor.
JNJ Independent Contractor.
Merck Independent Contractor.
Nuvalent Independent Contractor.
Nuvation Independent Contractor.
Onviv Independent Contractor.
Pfizer Independent Contractor.
Regeneron Independent Contractor.
Summit Independent Contractor.
Takeda Independent Contractor.
Taiho Independent Contractor.
Tempus Boehringer Ingelheim, Independent Contractor.
Boehringer Ingelheim, Travel.
Daiichi Sankyo Travel.
J. Tao, None.
P. M. Forde,
AstraZeneca Independent Contractor, ).
Bristol-Myers Squibb Independent Contractor, ).
Novartis ).
Corvus ).
Kyowa ).
Amgen Independent Contractor.
Daiichi Sankyo Independent Contractor.
Genentech Independent Contractor.
G1 Therapeutics Independent Contractor.
Iteos Independent Contractor.
Janssen Independent Contractor.
Merck Independent Contractor.
Surface Oncology Independent Contractor.
Mirati Independent Contractor.
Novartis Independent Contractor.
Sanofi Independent Contractor.
Polaris Other, DSMB member.
Flame Therapeutics Other, DSMB member.
J. C. Murray, None.
M. Sausen,
Labcorp Employment, Stock.
G. A. Meijer,
LabCorp ).
Delfi Diagnostics ).
G. Vink,
BMS ).
Merck ).
Servier ).
PGDx ).
Bayer ).
Sirtex ).
Pierre Fabre ).
Delfi Diagnostics ).
Natera ).
Nordic pharma ).
R. J. Fijneman,
Labcorp ).
Delfi Diagnostics ).
Solvias ).
Natera ).
V. E. Velculescu,
Delfi Diagnostics g., Board of Directors, non-salaried role), Independent Contractor, Stock.
LabCorp Other Intellectual Property.
Qiagen Other Intellectual Property.
Sysmex Other Intellectual Property.
Agios Other Intellectual Property.
Genzyme Other Intellectual Property.
Esoterix Other Intellectual Property.
Ventana Other Intellectual Property.
ManaT Bio Other Intellectual Property.
Danaher Independent Contractor.
Takeda Pharmaceutical Independent Contractor.
Viron Therapeutics Independent Contractor.
J. A. Phallen, None.
R. Scharpf,
Delfi Diagnostics Independent Contractor, Stock.
LabCorp Other Intellectual Property.
Qiagen Other Intellectual Property.
Sysmex Other Intellectual Property.
Agios Other Intellectual Property.
Genzyme Other Intellectual Property.
Esoterix Other Intellectual Property.
Ventana Other Intellectual Property.
ManaT Bio Other Intellectual Property.
V. Anagnostou,
AstraZeneca Independent Contractor, ).
Personal Genome Diagnostics/Labcorp ).
Bristol-Myers Squibb ).
Delfi Diagnostics ).
Guardant Health Independent Contractor, Other, honoraria.
Neogenomics Independent Contractor.
Foundation Medicine Other, honoraria.
Personal Genome Diagnostics Other, honoraria.
Roche Other, honoraria.