Abstract：
BACKGROUND: The performance of serum galactomannan (GM) for the diagnosis of invasive aspergillosis (IA) has been studied mainly in adults. Paediatric data are scarce and based on small and heterogeneous cohorts.
OBJECTIVE: To evaluate the performance of serum GM for the diagnosis of IA in a paediatric oncologic population at high risk of IA and to clarify the impact of antifungal prophylaxis on this test.
METHODS: We performed a retrospective study from January 2014 to December 2020 in the paediatric oncologic haematologic department of the University Hospital of Bordeaux. The diagnosis of IA was made using the recommendations of the EORTC and the MSGERC.
RESULTS: Among the 329 periods at high risk of IA in 222 patients, the prevalence of IA was 1.8% (3 proven and 3 probable IA). In the total population, the sensitivity, and the positive predictive value (PPV) were respectively 50% and 17.6%. Under antifungal prophylaxis, the sensitivity and PPV dropped, respectively, to 33.3% and 14.3%. In this group, the post-test probability of IA was 2% for a negative serum GM and only 14%.
CONCLUSIONS: In this large cohort of children at high risk of IA, the incidence of IA is low and the diagnostic performance of GM is poor, especially in the case of mould-active prophylaxis. Screening should be targeted rather than systematic and should be reserved for patients at highest risk for IA without mould-active prophylaxis. Combination with other tests such as Aspergillus PCR would increase the accuracy of GM in screening setting.
OBJECTIVE: To evaluate the performance of serum GM for the diagnosis of IA in a paediatric oncologic population at high risk of IA and to clarify the impact of antifungal prophylaxis on this test.
METHODS: We performed a retrospective study from January 2014 to December 2020 in the paediatric oncologic haematologic department of the University Hospital of Bordeaux. The diagnosis of IA was made using the recommendations of the EORTC and the MSGERC.
RESULTS: Among the 329 periods at high risk of IA in 222 patients, the prevalence of IA was 1.8% (3 proven and 3 probable IA). In the total population, the sensitivity, and the positive predictive value (PPV) were respectively 50% and 17.6%. Under antifungal prophylaxis, the sensitivity and PPV dropped, respectively, to 33.3% and 14.3%. In this group, the post-test probability of IA was 2% for a negative serum GM and only 14%.
CONCLUSIONS: In this large cohort of children at high risk of IA, the incidence of IA is low and the diagnostic performance of GM is poor, especially in the case of mould-active prophylaxis. Screening should be targeted rather than systematic and should be reserved for patients at highest risk for IA without mould-active prophylaxis. Combination with other tests such as Aspergillus PCR would increase the accuracy of GM in screening setting.
摘要：
背景：主要在成人中研究了血清半乳甘露聚糖（GM）用于诊断侵袭性曲霉病（IA）的性能。儿科数据很少，并且基于小型和异质队列。
目的：评估血清GM在IA高危儿科肿瘤人群中诊断IA的表现，并阐明抗真菌预防对该测试的影响。
方法：我们于2014年1月至2020年12月在波尔多大学医院儿科肿瘤血液科进行了一项回顾性研究。IA的诊断是根据EORTC和MSGERC的建议进行的。
结果：在222例IA高危患者的329个时期中，IA的患病率为1.8%(3例确诊和3例可能的IA).在总人口中,灵敏度,阳性预测值(PPV)分别为50%和17.6%。在抗真菌预防下，灵敏度和PPV下降，分别,分别为33.3%和14.3%。在这个群体中,血清GM阴性时IA的试验后概率为2%,仅为14%.
结论：在这一庞大的IA高危儿童队列中，IA的发病率低，GM的诊断性能差，特别是在霉菌活性预防的情况下。筛查应该是有针对性的而不是系统的，并且应该保留给没有霉菌预防的IA风险最高的患者。与其他测试如曲霉PCR的组合将提高GM在筛选设置中的准确性。
目的：评估血清GM在IA高危儿科肿瘤人群中诊断IA的表现，并阐明抗真菌预防对该测试的影响。
方法：我们于2014年1月至2020年12月在波尔多大学医院儿科肿瘤血液科进行了一项回顾性研究。IA的诊断是根据EORTC和MSGERC的建议进行的。
结果：在222例IA高危患者的329个时期中，IA的患病率为1.8%(3例确诊和3例可能的IA).在总人口中,灵敏度,阳性预测值(PPV)分别为50%和17.6%。在抗真菌预防下，灵敏度和PPV下降，分别,分别为33.3%和14.3%。在这个群体中,血清GM阴性时IA的试验后概率为2%,仅为14%.
结论：在这一庞大的IA高危儿童队列中，IA的发病率低，GM的诊断性能差，特别是在霉菌活性预防的情况下。筛查应该是有针对性的而不是系统的，并且应该保留给没有霉菌预防的IA风险最高的患者。与其他测试如曲霉PCR的组合将提高GM在筛选设置中的准确性。
