How Hard Can It Be?

How strange the world of clinical microbiology is when you compare the fields of mycology, parasitology, bacteriology and virology to each other. Such different possibilities, opportunities, limitations, and diagnostic challenges! The 3 month mortality rate of invasive aspergillosis, a disease mainly caused by Aspergillus fumigatus and seen in mainly patients with haematological malignancies, patients undergoing allogenic HSCT and patients in ICUs, may be as high as 60%, and therefore a quick and reliable diagnosis is mandatory to secure timely therapeutic intervention. But, - Aspergillus fumigatus happens to be ubiquitous, and contamination of patient samples, whether blood or airway samples, may always be a potential cause of false-positive test results, and one of the reasons why the use of PCR as a first line diagnostic tool in routine mycology labs is still limited. Antigen tests, such as the Galactomannan antigen test, which also allow quick diagnosis can also be false-positive, not only due to sample contamination, but also due to galactomannan residues in medical compounds, such as the widely applied antibiotic Tazocin (piperacillin-tazobactam), which means that patients who have been given this drug and who submit a blood sample for galactomannan testing may test slightly positive even in the absence of an Aspergillus infection.
These are only some classical examples. In the field of mycology, positive predictive values (PPV; i.e. what is the probability of disease given a positive test result) are sometimes unacceptably low, and the lower the prevalence of the disease, the lower the PPV. This means that you need a lot of experience and knowledge on pre-test-probability + data from clinical and diagnostic work-ups, including anamnestic details, to determine whether or not the patient should receive therapy, such as treatment with voriconazole, -  a relatively expensive drug.

Aspergillus fumigatus - the most common cause of invasive aspergillosis - on blood agar.

In the parasitology lab, however, things are quite different. Contamination of patient samples is rarely an issue, and in most cases not possible at all (disregarding DNA contamination of course). Specificity of microscopy is very often very high (close to 100%), which means that the PPV is very high even in cases where the disease is rare. Hence, if cysts of Giardia have been detected in your stool, it's due to the presence of the parasite in your body. It's a bit more tricky with PCR-based analyses, where the specificity does not rely on your ability to visually distinguish between e.g. Giardia and non-Giardia elements, but where it's all about designing oligos that anneal only to Giardia-DNA.
While in the mycology lab we struggle with low PPVs, one of the biggest challenges for me and my colleagues in the parasitology lab is to optimise the negative predictive value (NPV) of a faecal parasite diagnostic work-up - how can we rule out parasitic disease by cost-effectively putting together a panel of as few tests as possible?

There are many other differences. For instance, you can grow bacteria and fungi in the lab very easily, in fact, culture of bacteria and fungi is an essential diagnostic tool, which also allows you to submit the strain to antibiotic or antimycotic susceptibility testing and molecular characterisation/MALDI-TOF analysis in case you are not sure about the species ID. So, you have the strains right there in front of you, on agar plates, and they grow and grow, and you can keep them for as long as you like, - clean, non-contaminated strains on selective media.
You can't really do that with parasites, not nearly to the same extent and as easily, that is. For instance, you can culture Blastocystis directly from stool for sure (go here for the protocol), but only in the presence of bacteria (some of my colleagues do actually now and then manage to grow strains of Blastocystis in the absence of bacteria, they obtain what is called "axenic" cultures, but I believe that they cannot do it consistently and in limited time.). And it's a pity, since there is so much you can do when you have "clean" patient strains. Apart from susceptibility testing (which would actually be a bit difficult since Blastocystis is strictly anaerobic, so you can't really have it in microtiter plates or on RPMI plates on the table in front of you, but the strains could be challenged in the growth tubes), you can also extract DNA, and you would know that all the DNA that you extract from the isolate is from that particular strain, and not from bacterial contaminants. You can use the strain for production of antigens which can be used in ELISAs and used to generate mono- and polyclonal antibodies... Sequencing genomes of various subtypes would be a lot easier and quicker, and so on...

So, what appears obvious in one field of microbiology is not as obvious in another field, and vice versa. I wish Blastocystis was much easier to isolate. Dientamoeba too. Dientamoeba is probably as common as Blastocystis, and not rarely seen in co-infections. It is strange to contemplate that a parasite infecting hundreds of millions of people has not yet had its genome sequenced? We have no clue when it comes to effector proteins in Dientamoeba, and also for this parasite, what we know about its clinical significance relies mainly on epidemiological data.

There is no doubt that concerted efforts of experienced scientists should make it possible to develop appropriate and relevant culture protocols for these parasites. It does, however, require a lot of resources and time to get to know these common, but oh so fragile and reclusive little creatures...

Further reading:

Clark CG, & Diamond LS (2002). Methods for cultivation of luminal parasitic protists of clinical importance. Clinical microbiology reviews, 15 (3), 329-41 PMID: 12097242

Verweij PE, Kema GH, Zwaan B, & Melchers WJ (2012). Triazole fungicides and the selection of resistance to medical triazoles in the opportunistic mould Aspergillus fumigatus. Pest management science PMID: 23109245

Stensvold, C., Jørgensen, L., & Arendrup, M. (2012). Azole-Resistant Invasive Aspergillosis: Relationship to Agriculture Current Fungal Infection Reports, 6 (3), 178-191 DOI: 10.1007/s12281-012-0097-7

Maertens J, Theunissen K, Verhoef G, & Van Eldere J (2004). False-positive Aspergillus galactomannan antigen test results. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, 39 (2), 289-90 PMID: 15307045
 

Munasinghe VS, Stark D, & Ellis JT (2012). New advances in the in-vitro culture of Dientamoeba fragilis. Parasitology, 139 (7), 864-9 PMID: 22336222

The "Flagyl" Poll

For some reason the "Flagyl" poll in the right side bar of this blog was reset; the number of votes was approaching 100. The question was

"For those who have received metronidazole (Flagyl or Protostat) treatment for Blastocystis, please indicate whether you experienced no, transient or permanent improvement (or none of the above)"

The interesting thing is that there was a tie between "no improvement" and "transient improvement", and although this poll could have been heavily biased in numerous ways, it is still completely in line with our experience: Many patients report transient alleviation of symptoms, while others have no clinical benefit from Flagyl. Flagyl is an antibiotic targeting a wide range of bacteria and single-celled parasites. It is sometimes successful in terms of eradicating Dientamoeba fragilis, one of the most common parasites in the human intestine, and a parasite which may cause symptoms especially in children (we are currently conducting a randomised control clinical trial at Statens Serum Institut to explore clinical and microbiological effect of metronidazole treatment of children with D. fragilis).

Many people will get diagnosed with Blastocystis without knowing whether they might also be positive for D. fragilis (and vice versa). It is a complex situation, since both parasites are common, they are difficult to detect unless you use PCR or other specialised analyses, and in most labs they are not tested for on a routine basis. And if they happen to be part of the panel of organisms that is tested for, it may be so that insensitive methods are used for their detection, which means that only a fraction of the cases will be detected. So, this is a bit of a conundrum in itself!

So, it's not easy to know what causes the temporary alleviation in some patients. Is it due to parasite recrudescence? Is it due to parasite eradication with subsequent re-infection? And which parasite? Blastocystis? Dientamoeba? Any others? Or, is it due to perturbation of the intestinal flora in a "positive" direction, which is then gradually going back to normal? Placebo effect? There are possibly many more explanations...

However, deep sequencing of faecal samples pre- and post treatment of parasite-positive patients will probably answer many of our questions...

Literature:

Engsbro AL, Stensvold CR, Nielsen HV, & Bytzer P (2012). Treatment of Dientamoeba fragilis in Patients with Irritable Bowel Syndrome. The American journal of tropical medicine and hygiene PMID: 23091195

Engsbro AL, & Stensvold CR (2012). Blastocystis: to treat or not to treat ... But how? Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, 55 (10), 1431-2 PMID: 22893582

Blastocystis and Microbiomology

Speaking of metagenomics: The July 2012 issue of one of the most prestigious journals in the field of clinical microbiology, Clinical Microbiology and Infection (CMI – published by European Society of Clinical Microbiology and Infectious Diseases), focuses entirely on recent advances in metagenomics, including its implications on clinical microbiology. Several of the keynote speakers from the MetaHIT conference in Paris (March, 2012) have contributed with papers, including Rob Knight, Willem M. de Vos and Paul W. O’Toole. In his editorial, Didier Raoult, puts emphasis on mainly two things: 1) that we need to be patient with data obtained from studies using metagenomics, since currently some conclusions are pointing in different directions and data are still scarce, and 2) that metagenomic studies should be independent of financial support from commercial sources, such as the industry of antibiotics and probiotics.

Although it may be too early to make b/w inferences from data already published, I think that the pioneers in metagenomics teach us to re-think or at least modify several hypotheses about the role of intestinal microbes in gastrointestinal health and disease and pursue new and exciting trajectories. In this blog post I would like to highlight a few things that may be interesting to people who are not familiar with metagenomics, but who are interested in our gut flora and how it may impact our lives.

So, what is metagenomics? Well, only a few years ago, microbiologists were used to looking at one single organism at a time, when exploring the potential role of an organism in health and disease. They were dependent on isolating the organism, for instance by culture, in order to have sufficient material for molecular studies, and in order to avoid mix-up of data from contaminating organisms. However, the human intestinal microbiome (gut flora) is made up by a plethora of organisms, mainly prokaryotes (bacteria), but also to some extent eukaryotes (parasites and fungi), archaea and viruses. Metagenomics, facilitated by massive high-throughput parallel sequencing of nucleic acids extracted from human faecal samples, allows us to get a holistic picture of the entire gut flora of a person. I.e.: We move from examining one single species or organism at a time, to be analysing entire eco-systems. We get to know not only the composition of microbic species inhabiting our gut, but also how they impact our body physiology: Interestingly, Gosalbes et al. (2012) describe how the composition of the intestinal flora may differ significantly from person to person, but later shows that the active intestinal flora is fairly similar among healthy individuals. So, what’s the active flora? Briefly: while metagenomics analyses the DNA (16s rDNA) from the microbiome and hence provides us with data on the mere composition of microbes, including a quantification, metatranscriptomics looks at RNA communities by looking at 16S rRNA and mRNA transcripts. In this way, we get to know the function of the intestinal microbiota and can temporarily ignore the part of the microbial community that is in “stand-by” mode only. The collective genome of the intestinal microbes vastly surpasses the coding capacity of the human genome with more than 3 million genes - in comparison the human genome comprises 20,000-25,000 protein-coding genes.

So far, metagenomic studies have focused mainly on bacteria, and hence we know very little about how intestinal parasites directly or indirectly impact the remaining gut flora and the host, and, importantly, how the bacterial flora influences the presence and activity of parasites. This is due in part to methodological limitations, but mainly to the fact that the bacterial microbiome can be viewed as an organ of the human body (Baquero et al., 2012) taking care of vital and irreplaceable functions that the host is not otherwise capable of, ranging from energy and vitamin metabolism to epithelial barrier integrity and immune modulation (Salonen et al., 2012). Like any other organ, the microbiome has physiology and pathology, and the individual (and collective?) health might be damaged when its collective population structure is altered (Baquero et al, 2012). This is one of the reasons why studies of host-gut flora interactions have focused on bacteria.

One of the striking findings in metagenomic studies is that humans can be more or less successfully stratified into three enterotypes based on their intestinal flora (Arumugam et al., 2011):

We see that the three enterotypes are dominated by mainly three different types of bacteria (Bacteroides, Prevotelia and Ruminocoocus, respectively). However, as mentioned earlier, functional analysis (and probably a lot more sampling) is required to understand microbial communities. One of the interesting topics in this respect is how enterotypes correlate to different health/disease phenotypes; i.e. whether people with a certain gut flora are more prone to (a) certain type(s) of disease(s).There is preliminary evidence that variations in the microbiota are linked to diseases including bowel dysfunction and obesity.

In terms of parasites, I believe that in the near future we will see data revealing to which extent - if any - common intestinal micro-eukaryotes such as Blastocystis and Dientamoeba correlate with these enterotypes or other subsets of bacteria which will enable us to generate hypotheses on the interaction of micro-eukaryotes and the bacterial flora, which in turn may impact host physiology. I will expand a little more on this in an upcoming letter in Trends in Parasitology (article in press).

Interested in more: Why not have a look at Carl Zimmer's article in The New York Times about gut flora transplantation, or read about modulating the intestinal microbiota of older people to promote enhanced nutrition utilisation and to improve general health (O'Toole et al., 2012)... Also, have a look at my most recent blog post.

Literature:

O’Toole, P. (2012). Changes in the intestinal microbiota from adulthood through to old age Clinical Microbiology and Infection, 18, 44-46 DOI: 10.1111/j.1469-0691.2012.03867.x  

Gosalbes, M., Abellan, J., Durbán, A., Pérez-Cobas, A., Latorre, A., & Moya, A. (2012). Metagenomics of human microbiome: beyond 16s rDNA Clinical Microbiology and Infection, 18, 47-49 DOI: 10.1111/j.1469-0691.2012.03865.x  

Baquero, F., & Nombela, C. (2012). The microbiome as a human organ Clinical Microbiology and Infection, 18, 2-4 DOI: 10.1111/j.1469-0691.2012.03916.x  

Salonen, A., Salojärvi, J., Lahti, L., & de Vos, W. (2012). The adult intestinal core microbiota is determined by analysis depth and health status Clinical Microbiology and Infection, 18, 16-20 DOI: 10.1111/j.1469-0691.2012.03855.x

Arumugam, M., Raes, J., Pelletier, E., Le Paslier, D., Yamada, T., Mende, D., Fernandes, G., Tap, J., Bruls, T., Batto, J., Bertalan, M., Borruel, N., Casellas, F., Fernandez, L., Gautier, L., Hansen, T., Hattori, M., Hayashi, T., Kleerebezem, M., Kurokawa, K., Leclerc, M., Levenez, F., Manichanh, C., Nielsen, H., Nielsen, T., Pons, N., Poulain, J., Qin, J., Sicheritz-Ponten, T., Tims, S., Torrents, D., Ugarte, E., Zoetendal, E., JunWang, ., Guarner, F., Pedersen, O., de Vos, W., Brunak, S., Doré, J., Consortium, M., Weissenbach, J., Ehrlich, S., & Bork, P. (2011). Enterotypes of the human gut microbiome Nature, 474 (7353), 666-666 DOI: 10.1038/nature10187

Blastocystis Subtyping in Routine Microbiology Labs

When I speak to colleagues in and outside Europe and visit research portals and social media, including Facebook groups, I get the impression that Blastocystis subtyping is something that is still very rarely done, despite the fact that most clinical microbiologists and biologists acknowledge that subtypes may differ in terms of clinical significance and in other respects. We get new data on Blastocystis subtypes in various cohorts from time to time from research groups around the world, but all reports are characterised by relatively small sample sizes and subtyping methodology has not yet been standardised. This type of research is moreover challenged by the fact that Blastocystis is common in healthy individuals (i.e. people not seeing their GPs for gastrointestinal problems), and this makes it extremely difficult to identify the subtype distribution in the "background" population.

Although we recommend barcoding (see one of my previous posts) as the subtyping method of choice, there is no "official report" identifying the Blastocystis subtyping gold standard. Therefore, I'm currently setting up a lab project that is going to help us compare the most common methods used for subtyping in order to identify the one most suitable. I emphasise that the best method used for subtyping is not the PCR that should be used for diagnostic purposes, mostly due to the fact that PCRs for subtyping amplify 300-600 bp, which are much longer amplicons than the one we go for in diagnostic PCRs (typically 80-100 bp). We therefore recommend our novel TaqMan-based real-time PCR for initial diagnosis, or culture, which is inexpensive and relatively easy and provides you with a good source of cells for DNA extraction.
I hope that we will be able to come up with some robust data soon that will allow us to recommend the most suitable approach and hope to publish our results in a clinical microbiology journal of high impact, and I hope that this will prompt Blastocystis subtyping in many labs. Once this report has been published, I intend to upload a protocol here at the site where lab procedures for diagnosis and subtyping will be described in detail. Stay tuned!

Novel Blastocystis real-time PCR

Our new real-time PCR for Blastocystis has been published in JCM.

Check it out at

Development and evaluation of a genus-specific, probe-based, internal process controlled real-time PCR assay for sensitive and specific detection of Blastocystis.

or

http://www.ncbi.nlm.nih.gov/pubmed/22422846

and use it!