Flatus—A Rapid Diagnostic Tool?

Abstract & Commentary

Synopsis: Volatile organic compounds from fecal samples appear to have good sensitivity and specificity for many enteric pathogens and could lead to rapid diagnostic testing.

Source: Probert CS, et al. A novel method for rapidly diagnosing the causes of diarrhea. Gut. 2004;53:58-61.

From the Bristol Royal Infirmary in the United Kingdom comes a report of an unusual application of available technology that may actually have some practical applications.

Probert and colleagues took fresh or frozen fecal samples from 35 patients with identified enteric pathogens and compared them with 6 normal controls for the presence of volatile organic compounds (VOC). They used less than a gram of stool in 10 mL sealed "headspace" vials, then assayed the vapors produced with gas chromatography and mass spectroscopy. This methodology has been used previously to detect rotting vegetables.

Analysis indicated the major VOC to be phenols in normal subjects with significant amounts of indoles, terpenes, and hydrocarbons. Organic acids were present in all samples but did not appear to be useful markers. Benzaldehyde was also present.

When VOCs were analyzed for the 6 samples from patients with Clostridium difficile diarrhea, furans were found to be prevalent (6 out of 6 and accounted for 25-55% of the VOC detected), but 3-methylindole was present in only 1 of the 6. Identification of 2-furancarboxaldehyde without measurable 3-methylindole was found to have a sensitivity of 83% and a specificity of 97% for C difficile.

The 5 rotavirus samples all contained ethyl dodecanoate, whereas only 1 of the other 36 samples, that of an adenovirus, did. The sensitivity was 100%, and specificity calculated as 97%. Ethanol was also found and may be a product of enteric organisms but could not be eliminated as originating from ingestion.

Ammonia was found in 7 of 9 Norwalk virus infections, both subjects with an astrovirus, 2 of 5 with adenoviruses, 1 of 3 with giardiasis, 1 of 5 with rotavirus but in none of those with a pure bacterial etiology.

Five Campylobacter jejuni samples were also studied. Phenols, indoles, and volatile organic acids were abundant but not specific. What was unusual, however, was the absence of VOC of the terpenes/hydrocarbons group. This absence was true for only 3 of the other samples.

Probert et al speculate that VOC may be a useful tool in the diagnosis of specific pathogens and possibly bowel disease, as well.

Comment by Alan D. Tice, MD, FACP

The problem of identification of enteric pathogens is a major one, with increasing complexity and cost and a reporting time that makes the microbiology laboratory more relevant to epidemiologic investigations than acute patient care. In fact, most decisions about antimicrobial intervention are made on clinical grounds—with the possible help of a stain of the feces for leukocytes or a direct assay for rotavirus or Giardia antigens. Not only that, but the decision about antimicrobial therapy may be an important one, especially as antibiotics may be detrimental with C difficile and in E coli O157 cases but beneficial with other invasive pathogens.

The need for better, faster, and cheaper tests for enteric pathogens is obvious, especially in developing countries, where deaths from diarrheal illness are thought to take millions of lives each year. A bedside test would be ideal.

There are obvious questions to be raised in regard to extraneous factors that may influence VOC in flatus or "vapors." Diet, alcohol, antibiotics, and likely other factors may have major effects and would need to be controlled. The ideas, however, are novel and may possibly have practical applications.

To what extent an infectious diseases specialist can use his nose in addition to the usual clinical skills and astute judgment is debatable. I have heard some speak as if their sense of smell is comparable to the microextraction techniques described. To my knowledge, their competence has never been tested in any form with an evidence-based approach. It is, however, obvious to any clinician that patients do vary in odor and that some are relatively revealing, although they usually seem more related to poor dentition or personal hygiene. To what extent the odors of the VOC that are specific for pathogens and could be refined and identified brings forth a variety of applications for cost-effective clinical care, as well as potentially furthering the frontiers of aromatherapy, but even speculation in these areas is unwarranted at this point. In all likelihood, however, the pathogen-specific VOC of flatus in patients with diarrhea may not have a specific smell associated with them.

It is hoped this insight and the encouraging results will lead to applications in clinical care and rapid diagnostic methods.