This open-access article is distributed under

Creative Commons licence CC-BY-NC 4.0. IN PRACTICE

CLINICAL ALERT
Statement on analysis and interpretation of clinical
human gastrointestinal microbiome testing using next-
generation sequencing in South Africa
J M Wojno,1 MB BCh, MMed (Microbiology), FC Path (SA); E du Toit,2 BSc Hons, PhD; A Deffur,3 MB ChB, DTH, MMed (Internal Medicine),
Cert ID (SA) Phys, PhD; A Brink,2 MB BCh, MMed (Medical Microbiology)

1
Lancet Laboratories, Cape Town, South Africa
2
Division of Medical Microbiology, Department of Pathology and National Health Laboratory Service, Faculty of Health Sciences,
University of Cape Town, South Africa
3
Department of Medicine, Faculty of Health Sciences, University of Cape Town, South Africa

Corresponding author: J M Wojno ([email protected])

Advances in DNA sequencing technologies and computational tools over the past few years have led to vast improvements in the
metagenomic analysis of the human microbiota. While this has also significantly improved our understanding of the role of the host-
microbiome interaction in health and disease, the current clinical expectation is that testing, particularly of the gastrointestinal biome, can
be used to diagnose, manage and treat patients. The authors outline the available technologies and highlight current limitations of these
techniques to address this clinical demand. Through understanding the limitations of and need for more research and data collection, one
can improve the appropriate utilisation and interpretation, as well as the current rational clinical application of these techniques.

S Afr Med J 2021;111(3):203-205. https://doi.org/10.7196/SAMJ.2021.v111i3.15336

The evolution of humans has occurred in association with many The role of microbiome testing and how to use the results of these
forms of microbial life.[1] The human body is home to a variety of tests in an appropriate manner, also need to be established. The
microbes, including bacteria, archaea, fungi, parasites and viruses. immense challenge of microbiota assessments is the standardisation
This microbial community is referred to as the microbiota. The of microbiota testing and reporting, as there are a variety of
collective genome content of the microbiota or the microbial laboratory techniques. Even though new diagnostic approaches for
metagenome is known as the microbiome.[2] microbiome analysis are more advanced and accessible, these are
Microbiota can be found throughout the human body, which variable.
contains about as many microbial cells as there are cells in the human There is much interest in precision medicine and the tailoring
body.[3] of therapy according to individual microbiota profiles/functions
Technical advances have led to an exponential rise in microbiome based on accurate microbiota assessments. The future aim would
research and consequently to clinician and patient interest in the be to manipulate the microbiota to affect function and composition –
human microbiome, which has increased the clinical demand for thereby treating disease. The fundamental notion of primum
laboratory microbiome testing and diagnostics. It is anticipated that non nocere,[10] – ‘first, do no harm’ – should be inherent in such
these tests will provide enough information to diagnose and/or treatment. Any manipulation of the microbiota using strategies as
manage patients with chronic/complex conditions, where existing dramatic as faecal microbiota transplantation, or potentially simple
diagnostic and management tools might have been deficient. or ‘harmless’ interventions, such as probiotic administration, may
It has been well established that the microbiota play an important potentially have a deleterious effect. Therefore, a case for a donor
role in human biology and that the microbe-human interaction may stool bank in South Africa (SA) is proposed.[11] As an example, even
be significant in determining disease and health.[4] probiotics may have an effect on metabolic activities, cause systemic
Many factors, such as diet, age and medication, are known to infection, such as endocarditis in susceptible populations, and have
change the composition and function of gastrointestinal tract (GIT) inappropriate effects on host immune responses.[12]
microbiota.[5] As an example, urbanisation, which involves, e.g. changes
in diet, environmental exposures, antibiotics and caesarean section How are the human microbiota
delivery, has been associated with changes and/or an imbalance in assessed?
the gut microbiota and an increased risk of chronic diseases, such as The current standard ‘microbial analysis’ in clinical practice involves
asthma, inflammatory bowel disease and other systemic conditions, pathogen-specific diagnostic procedures, where clinicians suspect
e.g. obesity, atherosclerosis, type 2 diabetes and Alzheimer’s disease.[5-8] that a specific infectious aetiology causes a certain infectious clinical
Although there are clear associations between the composition and syndrome. These pathogen-specific tests, e.g. stool or sputum
functions of certain microbiota and specific metabolic/neurological/ microscopy, culture and sensitivity, and standard molecular tests,
autoimmune disorders, many of these relationships may not be causal. e.g. multiplex polymerase chain reaction (PCR), focus on the
Establishing causation, albeit more challenging, is important to differentiation and isolation of pathogens from normal flora. This
inform accurate diagnostic and safe management practices.[9] testing does not specifically address the detection of the ‘microbiome’

203 March 2021, Vol. 111, No. 3

in that particular environment, i.e. all of the uncultivable normal Once a context-dependent normal is established, can one more easily
commensals or transient colonisers or pathobionts. define imbalanced flora? The term dysbiosis is used in the literature
Assessment of the microbiome requires the use of robust protocols. to describe an unhealthy imbalanced ecosystem that results in
The standardisation of each step from sample collection to data disease or an unhealthy phenotype. This may be due to imbalance in
analysis is empirical, but still emerging. An important step is the physiological/metabolic functions of the microbes to the detriment
method used to profile the microbial community, which gives varying of the host.
results. For example, 16S rRNA next-generation sequencing uses Various methods have been used to assess dysbiosis. Some
primers that target one or more of the variable regions within the indicators include diversity and resilience.[15]
16S rRNA gene, giving a high-level, low-resolution overview of the
bacterial and archaeal community.[13] This method is well tested Diversity
and applicable to a wide range of sample types and study designs. Diversity in a microbial community refers to richness (number of taxa
It is also quick and cost effective. Whole metagenomic sequencing, present) and evenness (abundance of many microbial constituents).
however, provides a deeper insight into all the microbial genomes Tools have been developed to compare patient cohorts. These
present in a sample, including viral and eukaryotic, narrowing down include alpha diversity (variation of microbes in a single sample),
to strain-level resolution and allowing for functional assessment of e.g. the Shannon diversity index, which combines richness and
genes present.[13] This method is more expensive to perform and the diversity and measures the number of species and inequality between
analysis is more complex than the targeted gene approach mentioned species abundances. Beta diversity measures variation of microbial
above. communities between samples, e.g. the Bray-Curtis dissimilarity. It
The technical challenges include standardising of protocols for has been postulated that high diversity has generally been associated
specimen collection, laboratory processing, testing/analysis and with health. Although this principle has mainly referred to the GIT,
storage, reporting, interpretation, as well as various computational where a relative lack of diversity has been noted in many disease
methods. This may result in a variety of test results.[14] Taking into states and the skin where a low diversity is associated with, e.g. atopic
account the dynamic nature of the microbiota and the ability of the dermatitis,[17] diversity may not be the only marker of a healthy
composition to change quite rapidly, a single snapshot in addition to ecosystem.[18] This is because the principle of a low diversity only
the technical variation may produce a result that is very difficult to being present with disease, and a high diversity only being present
interpret. with health, is not always true.[18] Also, there are other anatomical
sites where an increased diversity approximating health may not
Normal v. abnormal microbiomes necessarily apply, e.g. the vaginal microbiome, where increased
Microbiota analysis to assess taxonomic content and abundance of diversity may be associated with bacterial vaginosis and therefore
organisms has been used to assess organisms at a particular site. local inflammation.[19]
The composition and abundance of organisms based on taxonomic The notion of a diverse community being healthy is in part due
grouping are assessed, e.g. at phylum, genus or species level. The to the concept of temporal stability (ability to withstand external
organisms may be grouped into operational taxonomic units, which perturbations and rapidly return to a healthy state). This is coined
are working definitions to classify groups of related organisms. ‘resistance’/‘resilience’, e.g. ‘colonisation resistance’, which is the
Once again, there is much variation in composition and abundance ability to withstand colonisation with a pathogen, thus potentially
of organisms in and between individuals. Because of temporal preventing disease.
variation/dynamics, the taxonomic content of the microbiota of It is pivotal to consider that there is no standard normal definition
various individuals may differ significantly and is highly personalised, and testing protocol to assess diversity and that it is dependent on
and thus a normal that is context dependent may be more appropriate context.[18] Lastly, and perhaps most importantly, we need to take
to consider. into account that the analytical variability/lack of standardisation of
Based on taxonomy, the assessment of a healthy core of organism testing methods has a major impact on the assessment of microbiota.
in people without disease has been proposed. Furthermore, the
possibility of finding organisms/biomarkers/disease signatures, Beyond the microbiome
where their presence is always indicative of a disease state, has also We have, however, noted that industrialisation has had an impact
been suggested.[15] on microbiota and the ‘disappearing microbe’ hypothesis, where
Defining context-specific ‘normal’ and ‘abnormal’ microbiota for certain microbes are no longer present in modern civilizations
a specific region may be important. To define ‘unhealthy/abnormal’, compared with primitive societies, with a loss of diversity as a result
one needs to describe normal, or normal core of organisms for a of industrialisation.[20]
specific area, i.e. normal vaginal, normal GIT. We know that each Researchers have argued that analysis based purely on taxonomic
region is vastly different in terms of taxonomic composition and content may not be appropriate, as it does not factor in the metabolic
function of the microbes[16] and thus each region has its own normal. activity/core functions of the population assessed.
How do we then characterise a normal healthy anatomical site- The concept of functional redundancy occurs when different
specific microbiome? A couple of important factors need to be taxonomic profiles lead to ecosystems with similar behaviour/
considered. These include physiological intra-individual variation (at function. Integration of both metabolomic and metagenomic data
the individual level) of the microbiota, i.e. circadian rhythm, sex, age, with standardised computational data analysis tools for the assessment
genetic factors, menstrual cycle and normal interindividual variation of the functions/metabolites of a particular community may be more
(at the population level) that may have an effect on the microbiota, accurate.[13] Multi-omic analysis is therefore a new approach, where
such as geographical location and race/ethnicity.[5] data sets of different omic groups are combined during analysis. The
If a patient sample is therefore tested at a certain time point, one different omic strategies employed during multi-omics are genome,
must consider that this snapshot of the microbiota at one point in proteome, transcriptome, epigenome, exome and microbiome, which
time may not be appropriate. may provide better clinical applications in the future.

204 March 2021, Vol. 111, No. 3

Microbiome analysis in routine diagnostic, treatment and preventive interventions and tools. A
patient management myriad of proof-of-concept or principle studies are imminent to
Advances in the microbiome testing pipelines, especially the unlock the multitude of functions and confirm cause and effect in
bioinformatics tools and applications in analysing the data, have been disease states before routine analysis is recommended.
valuable in finding associations with disease in humans and making
one aware of the importance of the microbiota.
Although the microbiome and health and disease are very topical Declaration. None.
and microbiome testing is becoming more accessible locally and Acknowledgements. We acknowledge Prof. Mark Nicol for motivation
internationally, it is important to note that testing of the microbiota with initiation of this article.
(microbiome) is currently not part of routine clinical diagnostics Author contributions. JMW and EdT contributed to the writing of the
and it may take some time before microbiome-based tests become article. AD and AB checked the article for clinical relevance and applicability.
routine practice. Funding. None.
The results of these tests should not be used for diagnostic Conflicts of interest. JMW and EdT have research interests in
purposes or to inform treatment and management decisions. If personalised medicine and clinical applicability of personalised GIT
testing is performed, the results should be used for informative microbiome reports.
purposes only. Although there is an abundance of literature on
the role of the microbiota in human health and disease, this field
is evolving and the use of these results and their clinical relevance 1. Rook G, Bäckhed F, Levin BR, et al. Evolution, human-microbe interactions, and life history plasticity.
need to be elucidated further. Associations with health and disease Lancet 2017;390(10093):521-530. https://doi.org/10.1016/S0140-6736(17)30566-4
2. Lederberg J, McCray AT. ‘Ome sweet omics’ ‒ a genealogical treasury of words. Scientist 2001;15(7):8.
are often merely associations and causality may/may not have been 3. Sender R, Fuchs S, Milo R. Revised estimates for the number of human and bacteria cells in the body.
established. PLoS Biol 2016;14(8). https://doi.org/10.1371/journal.pbio.1002533 
4. Proctor L, LoTempio J, Marquitz A, et al. A review of 10 years of human microbiome research activities
Standard medical protocols for diagnosis and treatment of patients at the US National Institutes of Health, fiscal years 2007 - 2016. Microbiome 2019;7(1):31. https://doi.
are advised as per local/national guidelines. org/10.1186/s40168-019-0620-y
5. Dominguez-Bello MG, Godoy-Vitorino F, Knight R, et al. Role of the microbiome in human
Another confounder is that very few data on the composition development. Gut 2019;68(6):1108-1114. https://doi.org/10.1136/gutjnl-2018-317503
6. Ley RE, Turnbaugh PJ, Klein S, et al. Microbial ecology: Human gut microbes associated with obesity.
and functions of the microbiota of SA patients are available as a Nature 2006;444(7122):1022-1023. https://doi.org/10.1038/4441022a
comparable reference range. What may be considered normal in 7. Larsen N, Vogensen FK, van den Berg FW, et al. Gut microbiota in human adults with type 2 diabetes
differs from non-diabetic adults. PLoS ONE 2010;5(2):e9085. https://doi.org/10.1371/journal.
Europe or the USA, may not necessarily be the case locally. Growing pone.0009085
our own SA microbiome database and approved biorepositories for 8. Jiang C, Li G, Huang P, et al. The gut microbiota and Alzheimer’s disease. J Alzheimer’s Dis
2017;58(1):1-15. https://doi.org/10.3233/JAD-161141
various clinical conditions may assist with research and answering 9. Neeraj K, Kasper S, Kasper DL. Moving beyond microbiome-wide associations to causal microbe
many questions in the future. This may potentially improve identification. Nature 2017;552(7684):244-247. https://doi.org/10.1038/nature25019
10. Sokol DK. ‘First do no harm’ revisited. BMJ 2013;347(7932):f6426. https://doi.org/10.1136/bmj.f6426
diagnostics and information for future clinical microbiota testing so 11. Fredericks E, Hoosien E, Brink A. The case for stool banks in South Africa. S Afr Med J 2019;109(8):546-547.
https://doi.org/10.7196/SAMJ.2019.v109i8.14169
that clinical relevance can be elucidated further. 12. Daliri EBM, Lee BH, Oh DH. Safety of probiotics in health and disease. In: Singh RB, Watson RR,
The authors encourage clinicians to consult experts in the field with Takahashi T, eds. The Role of Functional Food Security in Global Health. Cape Town: Academic Press,
2019:603-622.
regard to limitations of the available testing and to educate patients 13. Knight R, Vrbanac A, Taylor BC,  et al.  Best practices for analysing microbiomes.  Nat Rev
regarding the current heterogeneity and lack of diagnostic value. A Microbiol 2018;16(7):410-422. https://doi.org/10.1038/s41579-018-0029-9
14. Allaband C, McDonald D, Vázquez-Baeza Y, et al. Microbiome 101: Studying, analyzing, and
collaborative approach where universities, researchers, public and interpreting gut microbiome data for clinicians. Clin Gastroenterol Hepatol 2018;17(2):218-230.
private entities, such as laboratories, hospitals and biotechnology https://doi.org/10.1016/j.cgh.2018.09.017
15. Loyd-Price J, Abu-Ali G, Huttenhower C. The healthy human microbiome. Genome Med 2016;8(1):1-11.
experts, can all work together is required to achieve this goal. https://doi.org/10.1186/s13073-016-0307-y
16. Morgan XC, Segata N, Huttenhower C. Biodiversity and functional genomics in the human
In future, using microbiome methods for risk assessment to microbiome. Trends Genet 2013;29(1):51‐58. https://doi.org/10.1016/j.tig.2012.09.005
prevent or reduce disease and enhance/personalise treatment may 17. Kim JE, Kim HS. Review microbiome of the skin and gut in atopic dermatitis (AD): Understanding
the pathophysiology and finding novel management strategies. J Clin Med 2019;8(4):444. https://doi.
become a reality. To implement these into clinical practice, the org/10.3390/jcm8040444
limitations need to be addressed first. Personalised medicine is also 18. Johnson KVA, Burnet PWJ. Microbiome: Should we diversify from diversity? Gut Microbes
2016;7(6):455-458. https://doi.org/10.1080/19490976.2016.1241933
being explored for drug therapy to reduce side-effects or improve 19. Onderdonk AB, Delaney ML, Fichorova RN. The human microbiome during bacterial vaginosis.
efficacy of treatment and personalised nutrition.[21] Clin Microbiol Rev 2016;29(2):223-238. https://doi.org/10.1128/CMR.00075-15
20. Blaser MJ. The theory of disappearing microbiota and the epidemics of chronic diseases.
Immunology 2017;17(8):461-463. https://doi.org/10.1038/nri.2017.77
Conclusions 21. Zmora N, Zeevi D, Korem T, et al. Taking it personally: Personalized utilization of the human
microbiome in health and disease. Cell Host Microbe 2016;19(1):12-20. https://doi.org/10.1016/j.
Systematic analysis and cataloging of biome multi-omics in defined chom.2015.12.016

patient populations and understanding of the potential protective,

205 March 2021, Vol. 111, No. 3