Reducing the incidence of febrile neutropenia in cancer patients treated with chemotherapy with PointCheck: a portable non-invasive neutropenia analyzer

在美国，每年有65万癌症患者接受化疗。作为副作用， 他们的中性粒细胞计数（帮助抵抗感染的白色血细胞的数量）变低 （称为中性粒细胞减少症）并使其暴露于发热性中性粒细胞减少症（FN）， 严重的感染，需要昂贵的住院治疗，有时甚至死亡。如果 早期发现中性粒细胞减少症，可以通过及时提供FN来避免 预防性抗生素或生长刺激因子。目前，中性粒细胞测量需要 访问保健中心。这增加了感染风险（因为测量不频繁 并且在感染发生之前可能检测不到低计数）并降低化疗功效（如 为了安全，化疗水平保持在频率和剂量的下限）。我们的目标是克服 PointCheck是一种非侵入性中性粒细胞检测，可经常用于 当地卫生中心或家中，使医生能够尽量减少感染， 个性化化疗计划。我们技术的基础是无创光学成像， 甲襞毛细血管微循环。我们进行了一项试点临床研究， 护士使用PointCheck从46名化疗患者中获取数据。通过人工分析 训练数据集（N = 16名患者的38次测量），我们建立了一个标记的血流间隙数据集， 微循环可以被认为是中性粒细胞的代表，并显示其数量 可用于区分基线患者（> 500中性粒细胞/mm3）和IV级血小板减少患者 （<500/mm3）。然而，手动分析的需要使得患者无法使用该技术。 它的现状。在SBIR第一阶段的支持下，我们将（a）开发自动化 在我们的训练集上采用的手动数据处理工作流程，（B）在我们的训练集上测试这些算法。 获得的和未分析的测试数据库（N =来自30名患者的103个测量值）和（c）生成 额外的性能数据，包括进一步的中性粒细胞范围（> 500和<1，500 中性粒细胞/mm3），从而满足最小活产物的要求。的继承 在第二阶段，这些里程碑将使PointCheck能够用于多站点诊断验证 试验证实，在临床条件下，FN的管理可以改善与此 新模式

Every year in the US, 650,000 cancer patients are treated with chemotherapy. As a side effect, their neutrophil counts (number of white blood cells that aid in fighting infections) become low (referred to as neutropenia) and expose them to febrile neutropenia (FN) which are episodes of serious infection that entail costly hospitalization and sometimes death. If neutropenia is detected early, FN can be avoided through timely delivery of prophylactic antibiotics or growth stimulating factors. Currently, neutrophil measurements require visits to healthcare centers. This increases infection risk (as measurements are taken infrequently and low counts may not be detected before infections ensue) and reduces chemotherapy efficacy (as chemotherapy levels are kept at the lower end of frequency and dose for safety). We aim to overcome this limitation with PointCheck, a noninvasive neutrophil test that can be used frequently in local health centers or at home, enabling physicians to minimize infections and personalize chemotherapy planning. The basis for our technology is noninvasive optical imaging of microcirculation in nailfold capillaries. We conducted a pilot clinical study where trained nurses used PointCheck to acquire data from 46 chemotherapy patients. By manually analyzing a training dataset (N=38 measurements from 16 patients), we built a labeled dataset of flow gaps in the microcirculation that can be considered proxies of neutrophils and demonstrated their number can be used to distinguish baseline patients (>500 neutrophils/mm3) from grade IV neutropenic ones (<500/mm3). However, the need for manual analysis makes the technology inaccessible to patients on its current status. With support from SBIR Phase I, we will (a) develop algorithms that automatize the manual data processing workflow employed on our training set, (b) test those algorithms on our acquired and unanalyzed testing database (N=103 measurements from 30 patients) and (c) generate additional performance data including further intermediate neutrophil ranges (>500 and <1,500 neutrophils/mm3), thus fulfilling the requirements for a minimum viable product. The consecution of these milestones will enable, in Phase II, using PointCheck in a multi-site diagnostic validation trial to confirm, under clinical conditions, that the management of FN can be improved with this new paradigm.