Asthma breathomics—promising biomarkers in need of validation

Dear Editor, I read with interest the engaging review by Neerincx et al— Breathomics fromexhaledvolatile organic compounds (VOCs) in pediatric asthma. The results are encouraging suggesting good predictive accuracy for VOCprofiles in asthmadiagnosis. However the authors report on only those studies which they were aware of, limiting their scope to the last 10 years. Given that the first study of an exhaledVOC in asthmamaywell have been Olopade et al in 1997 we wanted to ascertain whether the review had indeed captured all the relevant literature. We conducted a systematic search using the following key words and MeSH terms—metabolomics, breathomics, exhaled breath, breath test, volatile organic compound, and asthma. Searcheswere conducted in PubMed, Medline Ovid, and Embase, with no date limits. Example search string (used in PubMed): ((“Breath Tests“[Mesh] OR “Exhalation“[Mesh] OR “exhaled“[All Fields] OR breath[All Fields]) AND (“Asthma“[Mesh] OR “asthma“[All Fields] OR “asthmatic“[All Fields]) AND (“Volatile Organic Compounds“ [Mesh] OR “Volatile Organic Compound*“[All Fields])) OR ((“asthma“ [MeSH Terms] OR “asthma“[All Fields] OR “asthmatic“[All Fields]) AND (Breathomic*[All Fields] OR (“metabolomics“[MeSH Terms] AND (“exhalation“[MeSH Terms] OR “exhalation“[All Fields] OR “exhaled“[All Fields] OR breath[All Fields] OR “breath tests“[MeSH Terms])))). The search process is summarized in the Prisma diagram below Figure 1. We identified 17 references to pediatric asthma breathomic studies. Six were abstracts, of which three were not published in full elsewhere; two presented identical data − early findings from a study by the review authors which is yet to be published; and one has since been published in full. The list of full studies identified by our search is similar but not identical to that of Neerincx et al. They included two papers examining pre-schoolwheeze (not asthma),whilewe identified a paper publishedmore than 10 years ago; this was a longitudinal study of asthmatic children living in a high pollution area (n = 26). This study reported largely non-significant results, with the exception of exhaled benzene which exhibited a moderate positive association with bothersome/severe asthma symptoms. The study was limited by the small number of breath samples obtained on symptom-free days (n = 6). An absence of significant findings is of course one reason why abstracts might not progress to full publication. While there are good reasons for excluding abstracts from a review—including the inability to assess the quality of studies and risk of bias—this does have the potential to give an unbalanced viewpoint; one which emphasizes positive associations and downplays null findings. Of the abstracts not subsequently published in full—Gahleitner et al do not report any results; while Brinkman et al report a cluster analysis identifying groups which differ significantly in clinical parameters. Wang et al found a correlation with night waking but not with asthma control; whereas Vijverberg et al report an area under the curve of between 0.71 and 0.97 for the identification of disease control. It would seem then that the review byNeerincx et al succeeded in capturing the majority of relevant literature; the additionally identified study adds little to their findings. The abstracts identified by our search present results largely relating to the ability of VOCs to differentiate between states of disease control, the results of whichwere conflicting. Overall we found little to either challenge or expand the findings of their review; the future of pediatric breathomics now lies in the ability to validate these findings in prospective cohort studies.