Breath Volatile Metabolites for the Diagnosis of Emerging Invasive Mold Infection

描述(由申请人提供)：这份重新提交的K23提案概述了Sophia Koo的五年发展计划，以实现作为一名以患者为导向的临床科学家的独立性，在侵入性霉菌感染(IMI)的新诊断模式的开发和评估方面拥有特别的专业知识。这项建议中概述的策略将使古医生能够在她护理IMI高危患者的临床背景和临床研究的基础上，确定目前可用于这些感染的诊断模式的局限性。 通过利用布里格姆妇女医院和Dana-Farber癌症研究所独特的临床和研究优势，Charles Stark Draper实验室的生物医学工程和生物信息学能力，以及哈佛公共卫生学院、东北大学和麻省理工学院的流行病学、生物统计学和分析化学培训机会。她将由哈佛医学院医学副教授林赛·巴登博士和哈佛大学公共卫生学院流行病学教授厄尔·弗朗西斯·库克博士共同指导。巴登博士是真菌诊断学专家，库克博士是研究设计和统计学习/数据挖掘方面的专家，包括分析高维生物标记物数据。顾博士还将与Daniel Kuritzkes博士、Raphael Dlin博士、Francisco Marty博士、John Perfect博士和John Wingard博士密切合作，他们将为拟议研究计划的执行提供职业指导和科学建议，她的合作者Paul Vouros博士、Terence Risby博士和James Comolli博士将为她的研究计划贡献他们在分析化学和呼气分析方面的专业知识。该研究计划将使用分析化学和统计监督学习方法，在体外和患者呼气中确定新出现的侵入性霉菌物种的挥发性有机化合物(VOC)代谢物模式，为新出现的IMI患者进行非侵入性、护理点呼气测试奠定基础。具体地说，本方案提出(1) 研究内容包括：(1)毛霉目、镰刀菌属和栅孢菌属的VOC代谢组的详细定义；(2)对新出现的IMI、临床和放射学结果提示有新出现的IMI的患者，以及由这些霉菌引起的呼吸道定植的患者呼吸中VOC代谢物的预期收集，以及VOC特征支持向量机分类算法的推导和验证，以区分新出现的IMI与其他真菌感染或非真菌肺部病变的患者。目前对这些感染的诊断方法是不充分的，而且往往是不可接受的病态，开发这种新的非侵入性诊断方法应该能够比目前可能的更早地诊断这些感染，对日益增长的新出现的IMI风险患者进行前瞻性监测，更合理地针对抗真菌治疗，并改善临床结果。

DESCRIPTION (provided by applicant): This resubmission K23 proposal outlines a five-year development plan for Sophia Koo to achieve independence as a patient-oriented clinician-scientist with particular expertise in the development and assessment of novel diagnostic modalities for invasive mold infections (IMI). The strategy outlined in this proposal will enable Dr. Koo to build upon her clinical background in caring for patients at risk for IMI and her clinicl research defining the limitations of currently available diagnostic modalities for these infections by taking advantage of the unique clinical and research strengths of Brigham and Women's Hospital and the Dana-Farber Cancer Institute, biomedical engineering and bioinformatics capabilities of the Charles Stark Draper Laboratory, and epidemiology, biostatistics, and analytical chemistry training opportunities at the Harvard School of Public Health, Northeastern University, and Massachusetts Institute of Technology. She will be co-mentored by Dr. Lindsey Baden, Associate Professor of Medicine at Harvard Medical School, and Dr. Earl Francis Cook, Professor of Epidemiology at the Harvard School of Public Health. Dr. Baden is an expert in fungal diagnostics, and Dr. Cook is an expert in study design and statistical learning/data mining, including analysis of highly dimensional biomarker data. Dr. Koo will also work closely with Drs. Daniel Kuritzkes, Raphael Dolin, Francisco Marty, John Perfect, and John Wingard, who will provide career guidance and scientific advice on the execution of the proposed research plan, and her collaborators Drs. Paul Vouros, Terence Risby, and James Comolli, who will contribute their expertise in analytical chemistry and breath analysis to her research plan. The research program will employ analytical chemistry and statistical supervised learning approaches to define volatile organic compound (VOC) metabolite patterns of emerging invasive mold species in vitro and in patient breath to lay the groundwork for a noninvasive, point-of-care breath test for patients with emerging IMI. Specifically, this program proposes (1) a detailed definition of the VOC metabolome of the Mucorales, Fusarium, and Scedosporium species that most commonly cause human IMI, and (2) the prospective collection of VOC metabolites in the breath of patients with emerging IMI, patients with clinical and radiographic findings suggestive of an emerging IMI, and patients with respiratory tract colonization by these molds, with derivation and validation of VOC feature support vector machine classification algorithms to discriminate patients with emerging IMI from patients with other fungal infections or non-fungal pulmonary processes. Current diagnostic modalities for these infections are inadequate and often unacceptably morbid, and development of this novel, noninvasive diagnostic approach should allow earlier diagnosis of these infections than currently possible, prospective surveillance in the growing population of patients at risk for emerging IMI, more rational targeting of antifungal therapy, and improved clinical outcomes.