User login
The intestinal microbiota, also commonly known as the “gut microbiome” is integral to human physiology and has wide-ranging effects on the development and function of the immune system, energy metabolism and nervous system activity. There is a lot of excitement around the potential of targeting the microbiome therapeutically to promote health and to prevent or treat medical conditions. Further, as DNA sequencing technologies and computational methods continue to improve (as reviewed by Rob Knight and colleagues in a prior editorial), there is significant interest in developing microbiome-based diagnostics for clinical applications.
The industry has recognized the consumer interest in microbiome-based diagnostics as an opportunity, and a number of commercial laboratories are marketing tests directly to patients. Physicians, particularly gastroenterologists, are increasingly being asked by their patients to help interpret such test reports; in some cases, the patient may even request a physician order to purchase the tests for insurance coverage or other reasons.
Earlier this year, AGA members had a robust discussion in the AGA Community about microbiome-based tests, requirements for physician authorization, and the clinical utility (if any) of the results of such tests. below.
Limitations of microbiome sequencing. Microbiome datasets have the same limitations as any other sample-dependent dataset. First and foremost, a single stool sample will tell you something about a person’s microbiome profile only at the time and location that the sample was collected. How the sample was collected and how it was stored may significantly impact the analysis. The analysis generally provides an overview of bacterial families and genera, but little information about the viruses, protozoa and fungi. Furthermore, stool analysis may not reflect well the microbiome composition at the mucosal surface in the intestine. As a result, a single analysis of an individual stool sample merely provides a snapshot of the fecal microbiome that is incomplete and extremely limited in what we can learn from it.
“Good” vs. “bad.” The reports resulting from microbiome-based tests often describe the patient’s microbiome profile in terms of how much “good” and “bad” bacteria are present. This kind of a classification framework represents a naïve and cartoonish view of the microbial world. Instead, it is important to appreciate microbial communities as functional networks, and that their functionality cannot be defined as a mere summation of individual microorganisms. Microbes, just like people, vary their behavior in accordance with the context that may be provided by the activity of other microbes and the host. Whether a particular species or strain is helpful or harmful depends on what other bacteria are present, their density, how they interact with each other (e.g., are they mutually beneficial or competitive?), and factors from the human host such as their diet or immune system activity. For example, Clostridioides difficile is a potential pathogen, yet it also naturally exists in the intestines of many people as a nonharmful, commensal species. Its pathogenic potential depends on the state of the other intestinal microbes and host factors, such as presence of anti-C. difficile toxin antibodies.
Importantly, microbiome tests, which generally provide only a low-resolution microbial community overview, are not designed for pathogen identification. That is best done with targeted diagnostics. Even then, as well illustrated by the C. difficileexample, diagnosis of an infection cannot be made on the basis of laboratory testing alone and requires clinical information.
Taxonomy vs. function. Current technology allows a fairly inexpensive characterization of most bacterial taxa (at family and genus levels). However, taxonomy is not easily translated into functional information. Different taxa of microbes may be able to execute the same chemical transformations. In contrast, functional information depends on the genes present and how much are these genes expressed. However, obtaining this kind of information is much more resource intensive. Measurements of metabolites may also provide very valuable functional information, but proper sample collection for metabolomics is much more difficult.
Interindividual variability. The consistent lesson we’ve learned from the microbiome literature is that there is not a single “healthy” microbiome profile. We have not identified a particular microbiome profile that is predictive of a particular disease, though many researchers are working to develop microbiome-based indices for diseases such as inflammatory bowel disease or obesity. Crowd-sourced studies such as the American Gut Project are working to expand and diversify microbiome datasets so that we can better understand the variability and begin to identify reproducible microbiome signatures. The microbiome data is extremely multidimensional and complex. Therefore, developing predictive patterns will likely require analyses of millions of samples linked to highly granular clinical metadata. Microbiome-based tests have potential to transform clinical care and become incorporated into the personalized medicine paradigm. However, we are at the very beginning of understanding what one’s microbiome profile means for their susceptibility to or progression of disease. As patients approach their health care providers with requests to order commercial microbiome-based tests or to help interpret a report, it is important to set the expectation that these tests are not well suited for diagnoses of infectious diseases or validated in specific diagnoses of any diseases. There are far more unknowns than knowns regarding the role of the microbiome and human health.
For those interested in learning more on this topic, I will be discussing it at the 2019 Gut Microbiota for Health World Summit with my colleague Diane Hoffmann, JD, MS, from the University of Maryland School of Law. The AGA Center for Gut Microbiome Research and Education’s scientific advisory board, on which Diane and I both serve, has also recognized the need for additional guidance. I would encourage my gastroenterology colleagues to continue sharing their experiences with microbiome-based tests through the AGA Community platform.
Recommended reading
- • Staley C, Kaiser T, Khoruts A. Clinician guide to microbiome testing. Dig Dis Sci. 2018 Sep 28. doi: 10.1007/s10620-018-5299-6.
- • Allaband C, McDonald D, Vazquez-Baeza Y, Minich JJ, Tripathi A, Brenner DA, et al. Microbiome 101: Studying, Analyzing, and Interpreting Gut Microbiome Data for Clinicians. Clin Gastroenterol Hepatol 2018. doi: 10.1016/j.cgh.2018.09.017. • Costello EK, Stagaman K, Dethlefsen L, Bohannan BJ, Relman DA. The application of ecological theory toward an understanding of the human microbiome. Science 2012. doi: 10.1126/science.1224203. Epub 2012 Jun 6.
Dr. Khoruts, of the University of Minnesota, is a member of the AGA Center for Gut Microbiome Research & Education scientific advisory board.
The intestinal microbiota, also commonly known as the “gut microbiome” is integral to human physiology and has wide-ranging effects on the development and function of the immune system, energy metabolism and nervous system activity. There is a lot of excitement around the potential of targeting the microbiome therapeutically to promote health and to prevent or treat medical conditions. Further, as DNA sequencing technologies and computational methods continue to improve (as reviewed by Rob Knight and colleagues in a prior editorial), there is significant interest in developing microbiome-based diagnostics for clinical applications.
The industry has recognized the consumer interest in microbiome-based diagnostics as an opportunity, and a number of commercial laboratories are marketing tests directly to patients. Physicians, particularly gastroenterologists, are increasingly being asked by their patients to help interpret such test reports; in some cases, the patient may even request a physician order to purchase the tests for insurance coverage or other reasons.
Earlier this year, AGA members had a robust discussion in the AGA Community about microbiome-based tests, requirements for physician authorization, and the clinical utility (if any) of the results of such tests. below.
Limitations of microbiome sequencing. Microbiome datasets have the same limitations as any other sample-dependent dataset. First and foremost, a single stool sample will tell you something about a person’s microbiome profile only at the time and location that the sample was collected. How the sample was collected and how it was stored may significantly impact the analysis. The analysis generally provides an overview of bacterial families and genera, but little information about the viruses, protozoa and fungi. Furthermore, stool analysis may not reflect well the microbiome composition at the mucosal surface in the intestine. As a result, a single analysis of an individual stool sample merely provides a snapshot of the fecal microbiome that is incomplete and extremely limited in what we can learn from it.
“Good” vs. “bad.” The reports resulting from microbiome-based tests often describe the patient’s microbiome profile in terms of how much “good” and “bad” bacteria are present. This kind of a classification framework represents a naïve and cartoonish view of the microbial world. Instead, it is important to appreciate microbial communities as functional networks, and that their functionality cannot be defined as a mere summation of individual microorganisms. Microbes, just like people, vary their behavior in accordance with the context that may be provided by the activity of other microbes and the host. Whether a particular species or strain is helpful or harmful depends on what other bacteria are present, their density, how they interact with each other (e.g., are they mutually beneficial or competitive?), and factors from the human host such as their diet or immune system activity. For example, Clostridioides difficile is a potential pathogen, yet it also naturally exists in the intestines of many people as a nonharmful, commensal species. Its pathogenic potential depends on the state of the other intestinal microbes and host factors, such as presence of anti-C. difficile toxin antibodies.
Importantly, microbiome tests, which generally provide only a low-resolution microbial community overview, are not designed for pathogen identification. That is best done with targeted diagnostics. Even then, as well illustrated by the C. difficileexample, diagnosis of an infection cannot be made on the basis of laboratory testing alone and requires clinical information.
Taxonomy vs. function. Current technology allows a fairly inexpensive characterization of most bacterial taxa (at family and genus levels). However, taxonomy is not easily translated into functional information. Different taxa of microbes may be able to execute the same chemical transformations. In contrast, functional information depends on the genes present and how much are these genes expressed. However, obtaining this kind of information is much more resource intensive. Measurements of metabolites may also provide very valuable functional information, but proper sample collection for metabolomics is much more difficult.
Interindividual variability. The consistent lesson we’ve learned from the microbiome literature is that there is not a single “healthy” microbiome profile. We have not identified a particular microbiome profile that is predictive of a particular disease, though many researchers are working to develop microbiome-based indices for diseases such as inflammatory bowel disease or obesity. Crowd-sourced studies such as the American Gut Project are working to expand and diversify microbiome datasets so that we can better understand the variability and begin to identify reproducible microbiome signatures. The microbiome data is extremely multidimensional and complex. Therefore, developing predictive patterns will likely require analyses of millions of samples linked to highly granular clinical metadata. Microbiome-based tests have potential to transform clinical care and become incorporated into the personalized medicine paradigm. However, we are at the very beginning of understanding what one’s microbiome profile means for their susceptibility to or progression of disease. As patients approach their health care providers with requests to order commercial microbiome-based tests or to help interpret a report, it is important to set the expectation that these tests are not well suited for diagnoses of infectious diseases or validated in specific diagnoses of any diseases. There are far more unknowns than knowns regarding the role of the microbiome and human health.
For those interested in learning more on this topic, I will be discussing it at the 2019 Gut Microbiota for Health World Summit with my colleague Diane Hoffmann, JD, MS, from the University of Maryland School of Law. The AGA Center for Gut Microbiome Research and Education’s scientific advisory board, on which Diane and I both serve, has also recognized the need for additional guidance. I would encourage my gastroenterology colleagues to continue sharing their experiences with microbiome-based tests through the AGA Community platform.
Recommended reading
- • Staley C, Kaiser T, Khoruts A. Clinician guide to microbiome testing. Dig Dis Sci. 2018 Sep 28. doi: 10.1007/s10620-018-5299-6.
- • Allaband C, McDonald D, Vazquez-Baeza Y, Minich JJ, Tripathi A, Brenner DA, et al. Microbiome 101: Studying, Analyzing, and Interpreting Gut Microbiome Data for Clinicians. Clin Gastroenterol Hepatol 2018. doi: 10.1016/j.cgh.2018.09.017. • Costello EK, Stagaman K, Dethlefsen L, Bohannan BJ, Relman DA. The application of ecological theory toward an understanding of the human microbiome. Science 2012. doi: 10.1126/science.1224203. Epub 2012 Jun 6.
Dr. Khoruts, of the University of Minnesota, is a member of the AGA Center for Gut Microbiome Research & Education scientific advisory board.
The intestinal microbiota, also commonly known as the “gut microbiome” is integral to human physiology and has wide-ranging effects on the development and function of the immune system, energy metabolism and nervous system activity. There is a lot of excitement around the potential of targeting the microbiome therapeutically to promote health and to prevent or treat medical conditions. Further, as DNA sequencing technologies and computational methods continue to improve (as reviewed by Rob Knight and colleagues in a prior editorial), there is significant interest in developing microbiome-based diagnostics for clinical applications.
The industry has recognized the consumer interest in microbiome-based diagnostics as an opportunity, and a number of commercial laboratories are marketing tests directly to patients. Physicians, particularly gastroenterologists, are increasingly being asked by their patients to help interpret such test reports; in some cases, the patient may even request a physician order to purchase the tests for insurance coverage or other reasons.
Earlier this year, AGA members had a robust discussion in the AGA Community about microbiome-based tests, requirements for physician authorization, and the clinical utility (if any) of the results of such tests. below.
Limitations of microbiome sequencing. Microbiome datasets have the same limitations as any other sample-dependent dataset. First and foremost, a single stool sample will tell you something about a person’s microbiome profile only at the time and location that the sample was collected. How the sample was collected and how it was stored may significantly impact the analysis. The analysis generally provides an overview of bacterial families and genera, but little information about the viruses, protozoa and fungi. Furthermore, stool analysis may not reflect well the microbiome composition at the mucosal surface in the intestine. As a result, a single analysis of an individual stool sample merely provides a snapshot of the fecal microbiome that is incomplete and extremely limited in what we can learn from it.
“Good” vs. “bad.” The reports resulting from microbiome-based tests often describe the patient’s microbiome profile in terms of how much “good” and “bad” bacteria are present. This kind of a classification framework represents a naïve and cartoonish view of the microbial world. Instead, it is important to appreciate microbial communities as functional networks, and that their functionality cannot be defined as a mere summation of individual microorganisms. Microbes, just like people, vary their behavior in accordance with the context that may be provided by the activity of other microbes and the host. Whether a particular species or strain is helpful or harmful depends on what other bacteria are present, their density, how they interact with each other (e.g., are they mutually beneficial or competitive?), and factors from the human host such as their diet or immune system activity. For example, Clostridioides difficile is a potential pathogen, yet it also naturally exists in the intestines of many people as a nonharmful, commensal species. Its pathogenic potential depends on the state of the other intestinal microbes and host factors, such as presence of anti-C. difficile toxin antibodies.
Importantly, microbiome tests, which generally provide only a low-resolution microbial community overview, are not designed for pathogen identification. That is best done with targeted diagnostics. Even then, as well illustrated by the C. difficileexample, diagnosis of an infection cannot be made on the basis of laboratory testing alone and requires clinical information.
Taxonomy vs. function. Current technology allows a fairly inexpensive characterization of most bacterial taxa (at family and genus levels). However, taxonomy is not easily translated into functional information. Different taxa of microbes may be able to execute the same chemical transformations. In contrast, functional information depends on the genes present and how much are these genes expressed. However, obtaining this kind of information is much more resource intensive. Measurements of metabolites may also provide very valuable functional information, but proper sample collection for metabolomics is much more difficult.
Interindividual variability. The consistent lesson we’ve learned from the microbiome literature is that there is not a single “healthy” microbiome profile. We have not identified a particular microbiome profile that is predictive of a particular disease, though many researchers are working to develop microbiome-based indices for diseases such as inflammatory bowel disease or obesity. Crowd-sourced studies such as the American Gut Project are working to expand and diversify microbiome datasets so that we can better understand the variability and begin to identify reproducible microbiome signatures. The microbiome data is extremely multidimensional and complex. Therefore, developing predictive patterns will likely require analyses of millions of samples linked to highly granular clinical metadata. Microbiome-based tests have potential to transform clinical care and become incorporated into the personalized medicine paradigm. However, we are at the very beginning of understanding what one’s microbiome profile means for their susceptibility to or progression of disease. As patients approach their health care providers with requests to order commercial microbiome-based tests or to help interpret a report, it is important to set the expectation that these tests are not well suited for diagnoses of infectious diseases or validated in specific diagnoses of any diseases. There are far more unknowns than knowns regarding the role of the microbiome and human health.
For those interested in learning more on this topic, I will be discussing it at the 2019 Gut Microbiota for Health World Summit with my colleague Diane Hoffmann, JD, MS, from the University of Maryland School of Law. The AGA Center for Gut Microbiome Research and Education’s scientific advisory board, on which Diane and I both serve, has also recognized the need for additional guidance. I would encourage my gastroenterology colleagues to continue sharing their experiences with microbiome-based tests through the AGA Community platform.
Recommended reading
- • Staley C, Kaiser T, Khoruts A. Clinician guide to microbiome testing. Dig Dis Sci. 2018 Sep 28. doi: 10.1007/s10620-018-5299-6.
- • Allaband C, McDonald D, Vazquez-Baeza Y, Minich JJ, Tripathi A, Brenner DA, et al. Microbiome 101: Studying, Analyzing, and Interpreting Gut Microbiome Data for Clinicians. Clin Gastroenterol Hepatol 2018. doi: 10.1016/j.cgh.2018.09.017. • Costello EK, Stagaman K, Dethlefsen L, Bohannan BJ, Relman DA. The application of ecological theory toward an understanding of the human microbiome. Science 2012. doi: 10.1126/science.1224203. Epub 2012 Jun 6.
Dr. Khoruts, of the University of Minnesota, is a member of the AGA Center for Gut Microbiome Research & Education scientific advisory board.