How do T-cells recognize foreign MHC molecules?


Question details: How do T-cells recognize foreign MHC molecules (and get activated thereafter), when the T-cells are positively selected in the thymus to recognize only own MHC (and foreign peptide, and get activated thereafter)?

T cell development occurs in the Thymus – Wikipedia. T cells bind pMHC (peptides bound to Major histocompatibility complex – Wikipedia (MHC)). In fact they can bind peptides presented both by MHC expressed by cells in the Thymus – Wikipedia of the body in which they develop, MHC restriction – Wikipedia, as well as by MHC expressed in genetically non-identical transplants, for e.g. during Allotransplantation – Wikipedia (1) *. The latter phenomenon is called alloreactivity.

As-Yet Unresolved Conundrums About T cell Development

The remarkable feature about the repertoire of B and T cells bearing unique B-cell receptor – Wikipedia (BCRs) and TCRs is they’re generated blind, i.e., in the absence of foreknowledge of antigens and antigen-derived peptides a person may encounter and need to respond to through their lifetime. T cell repertoire refers to the diversity of clonotypic T cells expressing unique TCRs, clonotypic meaning when a given T cell divides it creates a multitude of T cells bearing that same unique somatically generated TCR (Somatic recombination – Wikipedia).

T cells aren’t ‘positively selected in the thymus to recognize only own MHC (and foreign peptide)‘. To understand how that’s not even possible, consider someone who gets infected with flu. Thymic selection of T cells that ‘recognize only own MHC and foreign peptide‘ implies that to even have flu-specific T cells in the first place, the thymus should express flu antigens to positively select T cells expressing flu-specific TCRs, and so on ad nauseam for every one of the millions of peptides derived from the multitude of different kinds of entities a body might encounter and need to prevail over in the course of a lifetime. Crux is the thymus needs to select myriad T cell specificities during thymic T cell development even though the body can’t predict what antigen-derived peptides and antigens it would encounter in future. After all, the human adaptive immune system does have T (and B) cells that can specifically recognize and bind any number and variety of them. A conundrum indeed in that mature T cells bind specifically to both peptides (the norm in any immune response) and MHC molecules (the case in ‘direct’ response to allogeneic transplants) that they “couldn’t” have encountered during their development in the thymus.

  • How are TCR specificities selected during T cell development, as in the nature of the selecting thymic peptides. Cross-reactivity – Wikipedia is implicit in this process since clearly flu-derived peptides cannot have selected for a flu-specific T cell and so on.
  • Alloreactvity only adds to the conundrum since T cells also appear capable of binding MHC molecules they’ve obviously never previously encountered as happens with genetically mismatched transplants.

MHC Restriction Of T cells

Two principal models explain MHC restriction of T cells (2, 3, 4, 5, 6, 7).

Though not mutually exclusive, these models make different predictions. For the first one it doesn’t matter whether the T cell repertoire contains TCRs that bind pMHC or not while the second one requires TCRs biased to bind MHC, regardless of its class (I or II) or allele.

T cells That Make It Through Thymic Development Appear Wired to Bind MHC

Back in 1971, Niels Kaj Jerne – Wikipedia hypothesized ‘parallel evolution’ (8) of MHC and the then-undiscovered T-cell receptor – Wikipedia (TCR). After all, how else to explain alloreactivity other than by coevolution of TCR and MHC (9)?

As a recent illustrative example, a clever in vitro cellular model from 2016 demonstrates that TCRs may indeed be hard-wired to bind MHC (10; see figure below from 11).

This 2016 study thus concurs with many previous experimental (mostly mouse and some human) studies that found some Germline – Wikipedia amino acid residues on TCR alpha and beta chains to be crucial for binding MHC (4, 12, 13, 14, 15).

Cumulative data thus allows to infer that T cell development in the thymus may be mainly to ensure that only T cells with a ‘functional’ TCR get through developmental bottlenecks to be released into the ‘periphery’, i.e., a TCR capable of binding pMHC and delivering a modicum of signals downstream into the T cell, just enough, not too much nor too little, a la Goldilocks.

How To Explain T cell Alloreactivity (Ability of T cells to bind and respond to MHCs other than those that selected them in the thymus)

TCRs clearly bind both MHC molecules and the peptides they present. Two prevalent models to explain alloreactivity largely differ in which is more important, recognition of the peptides that allo MHC present, peptide-centric model, or the MHC molecules themselves, MHC-centric model (see figure below from (1).

Structural and some functional data from different experimental studies (16, 17, 18, 19) support either model.

One mouse TCR (17) was found to assume different conformations to accommodate binding to selecting versus novel MHC and in that study interfering with TCR’s ability to engage the peptide had little effect, i.e., support for the MHC-centric model. One mouse TCR was found to assume different conformations to accommodate binding to different peptides (16), i.e., support for the peptide-centric model. Meantime studies with human TCRs and HLA class I (18, 19) also support the MHC-centric model. Rather than one or the other, both approaches likely play their part in physiology.

* Note this answer deliberately avoids using ‘self’ and ‘non-self’/‘foreign’, mainstay words in immunology that obfuscate rather than clarify. In the age of Human microbiota – Wikipedia they’re also obviously unsuitable.


1. Boardman, Dominic A., et al. “What Is Direct Allorecognition?.” Current Transplantation Reports (2016): 1-9. What Is Direct Allorecognition?

2. Feng, Dan, et al. “Structural evidence for a germline-encoded T cell receptor–major histocompatibility complex interaction’codon’.” Nature immunology 8.9 (2007): 975-983.…

3. Dai, Shaodong, et al. “Crossreactive T Cells spotlight the germline rules for αβ T cell-receptor interactions with MHC molecules.” Immunity 28.3 (2008): 324-334.…

4. Garcia, K. Christopher, et al. “The molecular basis of TCR germline bias for MHC is surprisingly simple.” Nature immunology 10.2 (2009): 143-147.…

5. Garcia, K. Christopher. “Reconciling views on T cell receptor germline bias for MHC.” Trends in immunology 33.9 (2012): 429-436.…

6. Yin, Lei, et al. “T cells and their eons‐old obsession with MHC.” Immunological reviews 250.1 (2012): 49-60.…

7. Van Laethem, François, Anastasia N. Tikhonova, and Alfred Singer. “MHC restriction is imposed on a diverse T cell receptor repertoire by CD4 and CD8 co-receptors during thymic selection.” Trends in immunology 33.9 (2012): 437-441.…

8. Jerne, Niels Kaj. “The somatic generation of immune recognition.” European journal of immunology 1.1 (1971): 1-9.…

9. Felix, Nathan J., and Paul M. Allen. “Specificity of T-cell alloreactivity.” Nature Reviews Immunology 7.12 (2007): 942-953.

10. Parrish, Heather L., et al. “Functional evidence for TCR-intrinsic specificity for MHCII.” Proceedings of the National Academy of Sciences 113.11 (2016): 3000-3005.…

11. Krovi, Sai Harsha, and Laurent Gapin. “Revealing the TCR bias for MHC molecules.” Proceedings of the National Academy of Sciences 113.11 (2016): 2809-2811.…

12. Huseby, Eric S., et al. “How the T cell repertoire becomes peptide and MHC specific.” Cell 122.2 (2005): 247-260.…

13. Marrack, Philippa, et al. “Evolutionarily conserved amino acids in TCR V regions and MHC control their interaction.” Annual review of immunology 26 (2008): 171.…

14. Scott-Browne, James P., et al. “Germline-encoded amino acids in the αβ T cell receptor control thymic selection.” Nature 458.7241 (2009): 1043.…

15. Adams, Jarrett J., et al. “Structural interplay between germline interactions and adaptive recognition determines the bandwidth of TCR-peptide-MHC cross-reactivity.” Nature immunology 17.1 (2016): 87-94.…

16. Reiser, Jean-Baptiste, et al. “Crystal structure of a T cell receptor bound to an allogeneic MHC molecule.” Nature immunology 1.4 (2000): 291-297.…

17. Colf, Leremy A., et al. “How a single T cell receptor recognizes both self and foreign MHC.” Cell 129.1 (2007): 135-146.…

18. Archbold, Julia K., et al. “Alloreactivity between disparate cognate and allogeneic pMHC-I complexes is the result of highly focused, peptide-dependent structural mimicry.” Journal of Biological Chemistry 281.45 (2006): 34324-34332. Alloreactivity between Disparate Cognate and Allogeneic pMHC-I Complexes Is the Result of Highly Focused, Peptide-dependent Structural Mimicry

19. Macdonald, Whitney A., et al. “T cell allorecognition via molecular mimicry.” Immunity 31.6 (2009): 897-908.…

Why does rabies vaccine cause sarcoma in cats but not human?


What Is Feline Injection-Site Sarcoma (FISS)

  • First reported in the scientific literature in 1991 on cats originating from the US states of Pennsylvania, New Jersey and Maryland (1) but actually observed since at least 1987, feline injection-site sarcoma (FISS, Vaccine-associated sarcoma – Wikipedia) isn’t unique to inactivated rabies vaccine (2) but can also develop following FeLV (Feline leukemia virus – Wikipedia) vaccine (3), aluminum-adjuvanted vaccines (1), vaccines against feline panleukopenia virus (FPV), feline herpes virus-1 (FHV-1), feline calcivirus (FCV) or even just injections themselves in the absence of vaccines (4).
  • Multiple studies found no relationship between vaccine type, brand, nature (modified-live or inactivated) and FISS risk (5, 6, 7).
  • Injection-site sarcomas in cat occur even in the absence of vaccines (3, 8), for example following injections of glucocorticoids, antibiotics, anti-flea and painkillers (4).
  • Incidence of FISS ranged from 1.3 per 1000 (0.13%) in Canada from 1982 to 1993 (9) to 1 per 10000 (0.01%) in Canada and USA from 1998 to 2000 (10). In other words, relatively low with further reduction in recent years, perhaps as awareness of injection-related risk spread within the veterinary community and efforts were made to identify and mitigate risks.

FISS (Feline Injection-Site Sarcoma) Risk Factors

So what could trigger FISS?

  • Risk factors documented thus far include number of injections given in one site (more injections, higher the risk), route (interscapular, scruff of the neck, more risky) and temperature (cold more risky compared to room temperature) (5, see below from 11).
  • Adjuvanted vaccines may be more risky since histology and ultrastructural studies of FISS have shown adjuvants like aluminum concentrated within them (12, 13).
  • One risk factor may be genetic since risk is higher in siblings of cats with FISS (4). However, as of 2016 FISS genetic risk factors remain as yet undetermined (see below from 14),

‘Dr. Boston: Most affected cats are not purebreds; they’re domestic shorthairs from the pound. They’re spayed and neutered. No one knows where their littermates are. It’s difficult to show a genetic predisposition, but we presume it’s there. Some studies show that these cats are predisposed genetically for one of the cancer-promoter genes’

  • Research on linking FISS to tumor suppressor p53 – Wikipedia have so far been contradictory and hence inconclusive (15, 16, 17). This might have been due to different studies examining genetically disparate cat populations.

How To Minimize FISS Risk

Finally injection recommendations to minimize FISS occurrence include

  • Choosing non-adjuvanted, modified-live or recombinant vaccines over adjuvanted or killed vaccines (18).
  • Injecting only at recommended sites, hindlimbs, right forelimb and stringently avoiding interscapular (scruff of the neck) (see below from 11, 18).


1. Hendrick, Mattie J., et al. “Postvaccinal sarcomas in the cat: epidemiology and electron probe microanalytical identification of aluminum.” Cancer Research 52.19 (1992): 5391-5394. http://cancerres.aacrjournals.or…

2. Hendrick, M. J., and J. J. Brooks. “Postvaccinal sarcomas in the cat: histology and immunohistochemistry.” Veterinary Pathology 31 (1994): 126-126.…

3. Kass, Philip H., et al. “Epidemiologic evidence for a causal relation between vaccination and fibrosarcoma tumorigenesis in cats.” Journal of the American Veterinary Medical Association 203.3 (1993): 396-405.

4. Hartmann, Katrin, et al. “Feline injection-site sarcoma ABCD guidelines on prevention and management.” Journal of feline medicine and surgery 17.7 (2015): 606-613.…

5. Kass, Philip H., et al. “Multicenter case-control study of risk factors associated with development of vaccine-associated sarcomas in cats.” Journal of the American Veterinary Medical Association 223.9 (2003): 1283-1292.…

6. Wilcock, Brian, Anne Wilcock, and Katherine Bottoms. “Feline postvaccinal sarcoma: 20 years later.” The Canadian Veterinary Journal 53.4 (2012): 430.…

7. Srivastav, Anup, et al. “Comparative vaccine-specific and other injectable-specific risks of injection-site sarcomas in cats.” Journal of the American Veterinary Medical Association 241.5 (2012): 595-602.…

8. Martano, Marina, Emanuela Morello, and Paolo Buracco. “Feline injection-site sarcoma: past, present and future perspectives.” The Veterinary Journal 188.2 (2011): 136-141.…

9. Lester, Sally, Terri Clemett, and Alf Burt. “Vaccine site-associated sarcomas in cats: clinical experience and a laboratory review (1982-1993).” Journal of the American Animal Hospital Association 32.2 (1995): 91-95.

10. Gobar, Glenna M., and Philip H. Kass. “World wide web-based survey of vaccination practices, postvaccinal reactions, and vaccine site-associated sarcomas in cats.” Journal of the American Veterinary Medical Association 220.10 (2002): 1477-1482.…

11. Ladlow, Jane. “Injection Site-Associated Sarcoma in the Cat Treatment recommendations and results to date.” Journal of feline medicine and surgery 15.5 (2013): 409-418.…

12. Hendrick, M. J., et al. “Comparison of fibrosarcomas that developed at vaccination sites and at nonvaccination sites in cats: 239 cases (1991-1992).” Journal of the American Veterinary Medical Association 205.10 (1994): 1425-1429.

13. Madewell, B. R., et al. “Feline vaccine-associated fibrosarcoma: an ultrastructural study of 20 tumors (1996–1999).” Veterinary Pathology Online 38.2 (2001): 196-202. An Ultrastructural Study of 20 Tumors (1996-1999)

14. Boston, Sarah. “The Feline Sarcoma Controversy: Where Do We Stand?.”…

15. Banerji, Nilanjana, and Sagarika Kanjilal. “Somatic alterations of the p53 tumor suppressor gene in vaccine-associated feline sarcoma.” American journal of veterinary research 67.10 (2006): 1766-1772.

16. Banerji, Nilanjana, Vivek Kapur, and Sagarika Kanjilal. “Association of germ-line polymorphisms in the feline p53 gene with genetic predisposition to vaccine-associated feline sarcoma.” Journal of Heredity 98.5 (2007): 421-427. Association of Germ-line Polymorphisms in the Feline p53 Gene with Genetic Predisposition to Vaccine-Associated Feline Sarcoma

17. Mucha, D., et al. “Lack of association between p53 SNP and FISS in a cat population from Germany.” Veterinary and comparative oncology 12.2 (2014): 130-137.

18. Scherk, Margie A., et al. “2013 AAFP feline vaccination advisory panel report.” Journal of feline medicine and surgery 15.9 (2013): 785-808.…

How is it possible that a T Cell Receptor (TCR) recognises as few as 1-3 residues of the MHC-associated peptide?


, , , ,

Question details: In Basic Immunology, 5th Edition, it was mentioned that “each TCR recognizes as few as one to three residues of the MHC-associated peptide”. How is that possible while retaining specificity to the antigen? It seems to me that any single amino acid could be common to more than 1 peptide.

Short answer: Structural constraints on how many amino acid residues TCRs can bind on short peptide sequences include

  • Several of the peptide’s amino acid residues are already engaged in binding firmly to an MHC (Major histocompatibility complex) molecule creating the pMHC (peptide bound to MHC molecule).
  • At the same time the TCR also needs to bind some amino acid residues of the MHC molecule itself.

Different parts of the TCR bind peptide and MHC. The unique process of somatic nucleotide insertions and deletions makes the TCR’s Complementarity-determining region (CDR)3 region hypervariable and thus quite flexible and effective in binding a few conserved residues on a MHC-bound peptide. This feature also makes TCRs cross-reactive, i.e., capable of binding >1 unique pMHC.

Longer answer on theoretical basis for why it should be so and some data on how TCRs actually bind peptides

The adaptive immune system is predicated on the notion of anticipatory defense. CD4 and CD8 T cells express the alphabeta TCR, composed of two protein chains, alpha and beta. Each undergoes somatic gene rearrangement, specifically V(D)J recombination, i.e., nucleotide insertions and deletions at the V(D) J junctions in the Complementarity-determining region (CDR)3 regions of each chain of the TCR. This means each human body iterates from scratch its adaptive defense armamentarium, something that a 2015 immune parameter analysis of 210 healthy monozygotic twin pairs only confirms (1).

At first glance, this seems reasonable enough. After all, TCRs on CD4 and CD8 T cells have evolved to recognize and bind not whole protein molecules, ‘antigens’, but rather presented by MHC molecules, tiny pieces thereof, ‘peptides’, 12- to 20-mer (12-20 amino acids in length) in the case of CD4s, and 8- to 14-mer in the case of CD8s.

On second glance, this imposes a heavy burden on such a recognition system. Zooming in from larger structures to much smaller ones means the number of peptides presented to T cells by MHC class I and II molecules should be very large. Why should? Zooming in to such magnification simply exponentially increases the number of targets, i.e., peptides. If MHC molecules didn’t present as diverse of a peptide pool as possible from within a cell, missed pathogen-derived peptides become missed opportunities weakening the anticipatory potential of the adaptive immune system.

Given that TCRs bind short peptide sequences presented by MHC molecules, how could recognition be based on other than a handful of amino acid residues? Much of the rest of the peptide binds the MHC molecule. Peptide binding to MHC is itself a critical filtering event of great consequence in adaptive immunity. Antigen processing generates many peptides of varying lengths during protein digestion but only a handful succeed in making it past several bottlenecks to successfully and tightly bind MHC in such a way that they get presented on the cell surface, a feature called Immunodominance.

Why Most T cell Receptors (TCRs) Need To Be Inherently Cross-Reactive, Capable Of Binding >1 pMHC Complex

T cell receptors (TCRs) bind bits of both MHC and peptide. TCRs are hypothesized to have been evolutionarily selected to recognize and bind MHC. Thus, considering just the TCR engagement with the peptide cuts off too small a slice of the pie since outcome, TCR-mediated biochemical activation of T cells, depends on other critical factors,

  • How the peptide is bound within the MHC, which in turn determines which of its amino acid residues are available to make contact with the TCR and how optimally that could happen.
  • How the peptide is bound within the MHC also determines how optimally key anchor residues of the MHC molecule itself are available to bind the TCR.
  • How optimally can T cell co-receptors CD3 (immunology), CD4 (for CD4+ T cells)/CD8 (for CD8+ T cells) bind MHC at the same time.

Corollary of the fact that TCRs recognize and bind peptides presented by MHC molecules is the need for correspondingly large repertoire of specific T cells, one specific enough to recognize and bind each peptide processed and presented by each MHC molecule. However, if each TCR bound only one specific peptide, a logical extension of the Clonal selection theory (2, 3), each individual would theoretically need >10^15 T cells. Why? Because the 20 amino acid alphabet predicates the possible number of peptides that could bind to MHC molecules to be in the range of 10^15 (4). But 10^15 monospecific T cells necessary for optimal anticipatory defense entails a body weight of >500kgs (4), clearly and simply a physical impossibility, something that Don Mason already demonstrated in 1998 with his absurd mouse cartoon (see below from 5).

Given the constraint that there can only be far fewer T cells with unique TCRs compared to the potential diversity of possible unique pMHC complexes, no surprise TCRs tend to be cross-reactive, i.e., capable of binding >1 pMHC complex. An elegant 2014 experimental study found each of 5 different mouse and human TCRs even capable of binding >100 different peptides presented by one MHC molecule (6). Thus far crystal structures of ~120+ TCR- pMHC complexes have been published. One of the most striking observations of such crystal structures is the tremendous flexibility in how TCRs bind to pMHC (7). Rather than binding unique peptides, a given TCR can bind many but TCR binding’s supposed to be very specific. Cross-reactivity is antithetical to specificity. Can the two be reconciled? Yes, by the fact that the common thread linking cross-reactive peptides that bind a given TCR is the presence of conserved motifs, i.e., one, two, three or more conserved residues at specific TCR-binding positions (see quote, emphasis mine, and figure below from 6).

‘TCR cross-reactivity is not achieved by each receptor recognizing a large number of unrelated peptide epitopes but rather through greater tolerance for substitutions to peptide residues outside of the TCR interface, differences in residues that contact the MHC, and relatively conservative changes to the residues that contact the TCR CDR loops. The segregation of TCR recognition and MHC binding allows for TCRs to simultaneously accommodate needs for specificity and cross-reactivity.’

Thus, this example shows that even though peptides like 2A and MCC are very different in their sequences, both successfully bind the same TCR 2B4 because the process of somatic nucleotide insertions and deletions makes the region that primarily makes contact with a peptide, the TCR’s CDR3 region, hypervariable. This endows TCRs with the capacity to be quite flexible in accommodating different peptide sequences and also be able to bind firmly by contacting only a handful of conserved residues on any given MHC-bound peptide.

Human epidemiological studies reveal the implications of T cell recognition being the way it is. A relatively obscure set of data epitomize not only the extent of T cell cross-reactivity but also suggest that such functionality enables a vast, connected immunoprotective landscape against disparate entities ranging from bacteria to virus to cancer.

~85% of malignant melanocytes express an antigen called HERV-K-MEL (8, 9, 10), product of a pseudo-gene incorporated in the HERV-K env gene. HERV (Human Endogenous retrovirus) in turn are endogenous retroviruses incorporated into the human genome over millions of years. Acquired between 3 and 6 million years back, HERV-K are the latest family (11), making them the only HERVs still capable of replicating in the human population within the last few million years. HERV-K appears to be involved in several stages of melanoma formation (12, 13, 14).

Spontaneous melanoma regressions have occasionally been reported in the literature, suggesting effective anti-melanoma immune responses occur in nature. But what are the coordinates of such immunity? Taking a leaf out of William Coley and his Coley’s toxins, the European Organization for Research and Treatment of Cancer (EORTC) established the Febrile Infections and Melanoma (FEBIM) working group, tasked to explore how prior infectious diseases and vaccines influenced melanoma risk.

Their studies thus far suggest the Tuberculosis (TB) BCG vaccine, the Vaccinia vaccine against small pox and the 17D Yellow fever vaccine provide some degree of protection against melanoma (15, 16, 17).

What could possibly link such disparate characters as BCG, Vaccinia, Yellow fever and Melanoma and what connects this story to TCRs and peptides? Turns out each of these really disparate agents, a bacterium and two unrelated viruses, express peptides with high sequence homology to the melanoma HERV-K-MEL peptide (see below from 18).

BCG, vaccinia and yellow fever vaccines are of course expected to induce specific immune responses against themselves. However, given they express proteins with high sequence homology, vaccine-specific cytotoxic CD8+ and helper CD4+ T cells would also include those cross-reactive to melanoma HERV-K-MEL peptide. This could prevent melanoma development in those vaccinees who retain robust memory immune responses against this cross-reactive peptide.

Such phenomena may underlie the observation that certain live vaccines like BCG and measles can protect against unrelated pathogens and even reduce rates of all-cause mortality (19, 20). And as more microbiota-immunity interactions get mined, such examples that at present seem unanticipated will become more commonplace.


1. Brodin, Petter, et al. “Variation in the human immune system is largely driven by non-heritable influences.” Cell 160.1 (2015): 37-47.

2. Jerne, Niels K. “The natural-selection theory of antibody formation.” Proceedings of the National Academy of Sciences 41.11 (1955): 849-857.…

3. Jerne, Niels Kaj. “The somatic generation of immune recognition.” European journal of immunology 1.1 (1971): 1-9.

4. Sewell, Andrew K. “Why must T cells be cross-reactive?.” Nature Reviews Immunology 12.9 (2012): 669-677.…

5. Mason, Don. “A very high level of crossreactivity is an essential feature of the T-cell receptor.” Immunology today 19.9 (1998): 395-404.

6. Birnbaum, Michael E., et al. “Deconstructing the peptide-MHC specificity of T cell recognition.” Cell 157.5 (2014): 1073-1087.…

7. Rudolph, Markus G., Robyn L. Stanfield, and Ian A. Wilson. “How TCRs bind MHCs, peptides, and coreceptors.” Annu. Rev. Immunol. 24 (2006): 419-466.…

8. Kölmel, K. F., O. Gefeller, and B. Haferkamp. “Febrile infections and malignant melanoma: results of a case-control study.” Melanoma research 2.3 (1992): 207-212

9. Schiavetti, Francesca, et al. “A human endogenous retroviral sequence encoding an antigen recognized on melanoma by cytolytic T lymphocytes.” Cancer research 62.19 (2002): 5510-5516.…

10. Grange, John M., et al. “Can prior vaccinations against certain infections confer protection against developing melanoma?.” Medical Journal of Australia 191.9 (2009): 478.…

11. Sverdlov, Eugene D. “Retroviruses and primate evolution.” Bioessays 22.2 (2000): 161-171.

12. Muster, Thomas, et al. “An endogenous retrovirus derived from human melanoma cells.” Cancer research 63.24 (2003): 8735-8741.…

13. Serafino, A., et al. “The activation of human endogenous retrovirus K (HERV-K) is implicated in melanoma cell malignant transformation.” Experimental cell research 315.5 (2009): 849-862.…

14. Singh, Sarita, et al. “The role of human endogenous retroviruses in melanoma.” British Journal of Dermatology 161.6 (2009): 1225-1231.

15. Grange, John M., Bernd Krone, and John L. Stanford. “Immunotherapy for malignant melanoma–tracing Ariadne’s thread through the labyrinth.” European Journal of Cancer 45.13 (2009): 2266-2273.

16. Krone, Bernd, et al. “Protection against melanoma by vaccination with Bacille Calmette-Guerin (BCG) and/or vaccinia: an epidemiology-based hypothesis on the nature of a melanoma risk factor and its immunological control.” European Journal of Cancer 41.1 (2005): 104-117.

17. Mastrangelo, G., et al. “Does yellow fever 17D vaccine protect against melanoma?.” Vaccine 27.4 (2009): 588-591.

18. Cegolon, Luca, et al. “Human endogenous retroviruses and cancer prevention: evidence and prospects.” BMC cancer 13.1 (2013): 1. http://bmccancer.biomedcentral.c…

19. Goodridge, Helen S., et al. “Harnessing the beneficial heterologous effects of vaccination.” Nature Reviews Immunology (2016).

20. Muraille, Eric. “The Unspecific Side of Acquired Immunity Against Infectious Disease: Causes and Consequences.” Frontiers in microbiology 6 (2015).…

Are infections frequent during routine surgeries?


, , , , , , , ,

Post-Operative Infection Risk Is Multi-factorial. Patient’s Health Status Is Key Risk Factor

  • Routine surgery on who? Man or woman? Young or old? Person without or with conditions that predispose to post-operative infections, e.g., immunodeficiency, immunosuppression, diabetes, obesity, etc.?
  • What does ‘routine surgery’ entail? Superficial or invasive? Brief, less than a couple of hours, or prolonged, more than 5 or 6 hours? With local or general anesthesia? Brief or prolonged post-operative hospital stay? Device inserted during surgery or not? Surgical site closed (healing by primary intention) or left open (healing by secondary intention)?
  • Clearly, post-operative infection is the outcome of multiple factors (1). Not just the skill and rigor of the surgical team, skill in performing the surgery and rigor in maintaining sterility while doing it. Not just post-operative care that minimizes exposure to potential pathogens. The health status of the patient itself is a major factor in their post-operative infection risk. After all, harboring those sickest within a population makes hospitals a magnet for disease-causing microbes. The word Nosocomial means Hospital-acquired infection. Thus, language itself teaches us that likelihood of catching infections is heightened in hospitals, which is where most surgeries are performed.

More Frequent, Deadly Hospital-Acquired Infections Are Collateral Cost Of Widespread Human-Driven Antibiotic Resistance

Hospital-acquired infections have become more frequent in recent years for a few reasons.

  • More of us live longer though not necessarily in the best of health in old age.
  • Next, and perhaps most importantly, unnecessary antibiotic use has fueled global bacterial antibiotic resistance (2). Since hospitalized patients are those sickest among us, they’re also more likely to harbor and spread antibiotic-resistant bacteria. Such resistant bacteria can easily evade even the most rigorous control methods in even the wealthiest of countries (3, 4, 5, 6, 7, 8).
  • Finally, unrealistic expectations on the part of caretakers fuel unnecessary, prolonged and extensive medical assistance. As Atul Gawande has written extensively, the unrealistic attempt to evade death at all costs has taken root in countries like the US (9, 10). Even when such interventions are clearly futile, continuing with them has become part of a rote script, a script that includes more antibiotic Rx. Many elderly thus spend considerable time in nursing homes and long-term acute care centers. Hand in glove, cost-cutting measures in healthcare mean errors are more, not less, likely as fewer staff are mandated to provide futile care for those seriously ill (11). This makes infection transmissions between patients more, not less, likely. In turn such places become stable sources of antibiotic-resistant bacteria (11).

Beyond such technicalities is the self-evident difficulty in getting trustworthy data. After all, hospitals aren’t going to advertise their rates of post-operative infections, are they? Thus, these rates could vary vastly from one hospital to another and from country to country.

Some Numbers On Post-Operative Infections

Surgical Site Infections (SSI): Defined as an infection occurring within 30 days post-operation, surgical site infections (SSI) (see figure below from 12) are among the most common post-operative complications.

  • Between 1986 and 1996, the US CDC (Centers for Disease Control and Prevention) performed one of the 1st comprehensive SSI assessments (1). They assessed ~600000 operations of which ~2.6% (15523) developed SSI. 551/15523 (3.5%) of SSI patients died and 77% of those deaths were attributed to SSIs.
  • 2002 data suggested SSI were cause of >8000 deaths per year in the US (13).
  • A 2008 four country survey examined rates of healthcare-associated infections (HCAI) across acute hospitals in England, Wales, Northern Ireland and the Republic of Ireland and found SSI to be the 3rd most frequent nosocomial infection among hospital patients (14).
  • SSI were estimated to occur in 2.3% of cases based on 2005-2010 data from 30 hospitals in America, Asia, Africa and Europe (15).

Device-Associated Healthcare-Associated Infections (DA-HAIs)

  • As medicine increasingly incorporates complex technologies, approaches such as central-line catheters (16) to continuously deliver medicine directly into the bloodstream have mushroomed. No surprise, device-associated healthcare-associated infections (DA-HAIs) have become a major risk in the ICU.
  • A 2005 Canadian study found they’re a major cause of patient morbidity and mortality (17).
  • A 2002 to 2004 survey of 21069 patients in 55 ICUs of 46 hospitals across Argentina, Brazil, Colombia, India, Mexico, Morocco, Peru and Turkey found 3095 (14.7%) DA-HAIs (18).
  • Starting in 2006, the non-profit INICC (International Nosocomial Infection Control Consortium) published 5 pooled, multinational studies that suggest DA-HAIs in developing countries are 3 to 5 times higher compared to more developed economies (19).

Longer Answer: Read On If Interested

Specific Example Of How Infections Can Spread In Hospitals In Spite Of Best Efforts To Stop Them

Though not calling cards and therefore not advertised, every now and then an example comes along that offers us in great detail the process by which deadly, multi-drug resistant infections can take root and spread through a hospital these days, stubbornly evading and outwitting even the most determined and costly efforts to eliminate them.

Considered one of the world’s premier research hospitals, the case of the 2011 KPC (Klebsiella pneumoniae carbapenemase-producing K. pneumoniae) outbreak at the 243-bed US National Institutes of Health (NIH) National Institutes of Health Clinical Center (CC) offers a tragic example of how deadly infections can spread through a hospital in spite of the best precautions humans can think to devise (3, 4).

  • Patient one: On June 13, 2011, a 43-year old lung transplant patient with complications is transferred to the US NIH CC from a New York City hospital. On carefully checking her medical records, an alert infection control consultant notes she’s known to harbor a highly resistant ‘super bug’ called KPC (5). K. pneumoniae is a normal human gut inhabitant. Problem with KPC is it’s acquired additional antibiotic resistance rendering it a multi-drug resistant ‘super bug’, leaving only two less-than-optimal antibiotic choices (colistin, tigecycline) to treat it. Never having dealt with a KPC-harboring patient previously, the NIH CC takes obvious and necessary precautions, placing this patient in strict isolation within the ICU. This meant everyone entering her room donned a new protective gown and gloves, and even rigorously washed their hands after. Even her medical equipment gets specially decontaminated (6). Meantime, all other ICU patients also get their throats and groins tested regularly to track if KPC’s spread from patient one (6, 7). At first, all is well. The patient spends 24 hours in the ICU, is transferred to a private room, briefly returns to the ICU on June 29, then recovers and is discharged on 15 July, 2011.
  • Patient #2: Weeks after patient one leaves, on August 5, 2011, a 34-year old male ICU cancer patient shows KPC infection. No overlap between the two patients, either in ICU presence or healthcare staff who cared for them. Suggests KPC’s somehow stably present in the ICU.
  • Patient #3: On August 15, 2011, a 27-year old female patient shows KPC infection.

And thus a dreadful, seemingly inexorable process unfolds over the next four months. Starting in August other patients start acquiring KCP at the rate of ~1 per week (8) and eventually, a total of 18 patients come down with KPC and 7 die (see below from 3, 4). KPC-positive patients weren’t just in the ICU but also among non-ICU, meaning KPC had somehow escaped out of the ICU. Since patients at the NIH CC are usually seriously ill and only there by invitation to participate in a clinical trial, some had recently undergone chemotherapy, some had Leaky gut syndrome, some had been on Medical ventilator, and some had been on central-line catheters. In other words, all seriously sick and most with medical interventions that increase chances of bloodstream infections.

While this unfolds, the hospital takes unprecedented measures to root out and eliminate KPC from its ICU.

  • Patients and staff are grouped to eliminate any scope that staff caring for someone with KPC comes in contact with someone who doesn’t.
  • Every patient is repeatedly checked for KPC by sampling multiple sites on their bodies.
  • Entire rooms are fumigated with peroxide.
  • Plumbing lines are ripped out and replaced.
  • Finally, even the ICU is rebuilt.

Yet none of this seems to help.

How did patient one’s KPC spread in spite of such efforts? Genetic sleuthing reveals strains isolated from all these subsequent patients resemble that found in patient one and the transmission was anything but straightforward. Mutations it acquired as it spread allowed the geneticists to decipher its transmission path (5). In fact, genetic sequencing suggested KPC spread from patient one’s in three clusters (7),

  • From patient one’s throat to patient three who, while infected and asymptomatic, spread it to patients five and two.
  • From patient one’s lung to patient four from whom it spread to every other patient except one.
  • From patient one’s lung to patient eight and not spreading beyond.

Patients one and four were in different wards and never in the ICU at the same time, suggesting a silent carrier linked them, one who remained undiscovered. KPC was spreading stealthily in ways that wouldn’t be picked up by patient throat and groin cultures, i.e., by standard surveillance practice. Then, just as mysteriously as it started, KPC stops spreading in December 2011. Stops spreading but ominously, doesn’t disappear.

  • In April 2012, a young Minnesota man with severe graft-versus-host disease and Pseudomonas aeruginosa-associated pneumonia is admitted to the CC.
  • Shortly after, he tests positive for KPC. Genetic sequencing shows it’s the same strain first isolated there in June 2011.
  • On September 7, 2012, this young man dies in the isolation ward.
  • Several days later NIH staff swab a handrail outside his room and culture the same KPC from it.

Funded by the US government, evidently cost is no bar at a place like the NIH CC when it comes to stopping spread of infections. Yet this example shows how nearly impossible it is to do so even there (see below from 5, emphasis mine).

‘What was unusual about the Clinical Center’s experience with KPC klebsiella was not that it had an outbreak but that it quickly identified it and responded with such vigor. According to epidemiologists, in many other hospitals the patients would simply have died of an unspecified bloodstream infection, without anyone ever knowing the precise cause of their illness or how the infection had spread.

That will likely change. DNA sequencing is rapidly becoming more affordable. As a result, all hospitals will eventually have access to the tools that now exist only at NIH and other very specialized hospitals. However, very few will be able to afford to take the steps NIH did to contain the outbreak.’

And that’s not all. Decades of unnecessary antibiotic use have made outbreaks of such deadly antibiotic-resistant ‘superbugs’ more, not less, inevitable. Rampant antibiotic use in global industrial livestock production means that antibiotics are now everywhere in our environment, having leached into soil and entered waterways (2), thereby applying antibiotic resistance-selection pressures on all manner of microbes everywhere, not just on those associated with humans. Since many antibiotic resistance mechanisms can be horizontally transferred between bacteria, stopping unnecessary human antibiotic consumption alone may not minimize chances of such outbreaks.

‘Superbugs’ like KPC are now spreading faster than our capacity to control them. See below from 20 the rate and extent of KPC spread across the US since just 1999. For example, 2010-2011 surveys in Maryland, the state in which the NIH CC is located, found that ~80% of hospitals in that state had identified at least one case of carbepenem-resistant enterobacteriaceae like KPC (5). As things stand, this means risk of post-operative infections is now counter-intuitively higher, especially among the elderly and those with pre-existing conditions.


1. Mangram, Alicia J., et al. “Guideline for prevention of surgical site infection, 1999.” American journal of infection control 27.2 (1999): 97-134.…

2. Tirumalai Kamala’s answer to If we know that overusing antibiotics will cause resistant bacteria, why do we still give out so much of it? Especially in some parts of the world?

3. Snitkin, Evan S., et al. “Tracking a hospital outbreak of carbapenem-resistant Klebsiella pneumoniae with whole-genome sequencing.” Science translational medicine 4.148 (2012): 148ra116-148ra116. Tracking a Hospital Outbreak of Carbapenem-Resistant Klebsiella pneumoniae with Whole-Genome Sequencing

4. Lau, A., et al. “Laboratory response to a KPC outbreak at the NIH Clinical Center. Abstr. 112th Gen.” Meet. Am. Soc. Microbiol (2012): 16-19.…

5. Washingtonian, John Buntin, June 4, 2013. Outbreak at NIH | Washingtonian

6. Promed, Sep 17, 2012. ProMED-mail post

7. Bethesda magazine, Bara Vaida, Jan-Feb, 2013. The KPC Killer

8. Wired, Maryn McKenna, Aug 24, 2012. The ‘NIH Superbug’: This Is Happening Every Day

9. New Yorker, Atul Gawande, May 11, 2015. America’s Epidemic of Unnecessary Care

10. Being Mortal: Medicine and What Matters in the End: Atul Gawande: 9780805095159: Books

11. Scientific American, Judy Stone, Aug 24, 2012. The NIH Superbug Story a Missing Piece

12. Young, Pang Y., and Rachel G. Khadaroo. “Surgical site infections.” Surgical Clinics of North America 94.6 (2014): 1245-1264.…

13. Klevens, R. Monina, et al. “Estimating health care-associated infections and deaths in US hospitals, 2002.” Public health reports (2007): 160-166.…

14. Smyth, E. T. M., et al. “Four country healthcare associated infection prevalence survey 2006: overview of the results.” Journal of Hospital Infection 69.3 (2008): 230-248.…

15. Rosenthal, Victor D., et al. “Surgical site infections, International Nosocomial Infection Control Consortium (INICC) report, data summary of 30 countries, 2005–2010.” Infection Control & Hospital Epidemiology 34.06 (2013): 597-604.

16. Vox, Sarah Kliff, July 9, 2015. Do no harm: There’s an infection hospitals can nearly always prevent. Why don’t they?

17. Laupland, Kevin B., et al. “One-year mortality of bloodstream infection-associated sepsis and septic shock among patients presenting to a regional critical care system.” Intensive care medicine 31.2 (2005): 213-219.…

18. Rosenthal, Victor D., et al. “Device-associated nosocomial infections in 55 intensive care units of 8 developing countries.” Annals of internal medicine 145.8 (2006): 582-591.…

19. Rosenthal, V. D., et al. “International Nosocomial Infection Control Consortium (INICC) report, data summary of 43 countries for 2007-2012. Device-associated module.” American journal of infection control 42.9 (2014): 942.…


What are the similarities of people having the same blood group?



People with the same blood group have similar risk for various diseases, specifically certain cancers, cardiovascular disease, and increased susceptibility and more adverse outcome to some infectious/communicable diseases.

Blood group is typically defined by presence or absence of certain antigens on the surface of RBCs (Red blood cell). Of the ~35 different current blood group systems, the ABO blood group system, the focus of this answer, is considered the most important (see below from 1, 2, 3), especially for transfusion medicine.

Ironies About The Human ABO Blood Group System

  • Though Karl Landsteiner discovered them all the way back in 1900, hardly anything is known about their function.
  • However, function they must have not only because that’s nature’s way but also because they are widely expressed, in endodermal origin tissues of more ancient species such as amphibians and reptiles while also being expressed in ecto- and mesodermal tissues of more recently evolved species such as rodents and primates. Thus they’re expressed on epithelium, sensory neurons, platelets and blood vessels.
  • Ironically for molecules whose imprint on the public imagination is precisely their blood cell expression and ensuing need to match blood types during transfusion, their expression on human RBCs is an exception (see below from 4, 5).

Human Global ABO Distribution (6)

  • O is most common globally.
  • A is mainly found in North and Central Europe, rarer in Asia.
  • B is quite frequent in Central Asia but almost absent among Amerindians, who are almost exclusively O.
  • Clearly, the ABO blood group system exemplifies the fact that human blood group antigens are under active, intensive evolutionary selection pressure.

Human ABO Blood Group System & Disease Risk

  • Almost on the heels of the 1900 ABO blood group discovery, people started trotting out associations of different blood types with various human traits, ranging from personality to appropriate diets, the so-called blood group diet being merely one of the latest such fads. Thus far, being scientifically unsubstantiated is the only clear trait linking such tall claims (7).
  • OTOH, for many decades little research was done on differential risk for many diseases with different blood groups but in recent years a steady drip of data supporting such associations has emerged.
  • ABO blood groups appear to influence susceptibility to pancreatic and gastric cancers, cardiovascular disease and some infectious/communicable diseases (8; see below from 9) with type O tending to have lower risk.


  • Several recent epidemiological studies (10, 11, 12, 13, 14) suggest O blood type has lower risk of pancreatic cancer.
  • Since 1953 (15), epidemiological studies (16, 17) have found blood type A has increased risk of gastric cancer.

Cardiovascular Disease

  • Confirmed by 3 Meta-analysis (18, 19, 20), non-O blood types have increased risk of VTE (Venous thrombosis), apparently because circulating Von Willebrand factor is studded with ABH structures in A, B and AB blood types, which in turn increases thrombotic (clotting) risk (see below from 1).
  • OTOH, type O tends to have substantially lower levels of circulating von Willebrand factor, which further reduces clotting risk.
  • Increased VTE risk is especially well-documented for non-O blood types (21, 22, 23).
  • O blood type also has less risk of IS (Stroke), MI (Myocardial infarction) and PAD (Peripheral artery disease) (24).

Infectious/Communicable diseases

  • The specific O sub-type, O Lewis b, is associated with
    • Helicobacter pylori triggered-peptic ulcer, apparently because certain H. pylori strains can bind O lewis b antigen much more strongly compared to A lewis b (25).
    • Type O are also more susceptible to severe infections with Vibrio cholerae and Escherichia coli (8), and to norovirus-associated acute gastroenteritis (26).
  • OTOH, type O blood has better outcome and less severe symptoms from malaria, which provides a compelling example of microbial selection pressure on evolution of blood group antigens (27; see below from 28, 29). Protective, i.e., less severe, outcome for type O has been shown both experimentally (30) as well as through GWAS (Genome-wide association study) (31, 32).

Type O may do better because malaria-infected RBCs express novel proteins such as PfEMP-1 on their surface. In turn, in those with type A or B, such molecules can bind A and B antigens on platelets and blood vessels, setting off an aggregation cascade (see below from 27).


1. Franchini, Massimo, et al. “ABO blood group, hypercoagulability, and cardiovascular and cancer risk.” Critical reviews in clinical laboratory sciences 49.4 (2012): 137-149.…

2. Cid, Emili, et al. ABO in the Context of Blood Transfusion and Beyond. INTECH Open Access Publisher, 2012.…

3. Quinn, J. G., et al. “Blood: tests used to assess the physiological and immunological properties of blood.” Advances in physiology education 40.2 (2016): 165-175. Advances in Physiology Education

4. Oriol, Rafael, Jacques Pendu, and Rosella Mollicone. “Genetics of ABO, H, Lewis, X and related antigens.” Vox sanguinis 51.3 (1986): 161-171.

5. Eastlund, T. “The histo‐blood group ABO system and tissue transplantation.” Transfusion 38.10 (1998): 975-988.

6. Storry, J. R., and Martin L. Olsson. “The ABO blood group system revisited: a review and update.” Immunohematology 25.2 (2009): 48-59.…

7. Daniels, G. “The myths of blood groups.” ISBT Science Series 9.1 (2014): 131-135.

8. Franchini, Massimo, and Carlo Bonfanti. “Evolutionary aspects of ABO blood group in humans.” Clinica Chimica Acta 444 (2015): 66-71.

9. Yamamoto, Fumiichiro, et al. “ABO research in the modern era of genomics.” Transfusion medicine reviews 26.2 (2012): 103-118.

10. Wolpin, Brian M., et al. “ABO blood group and the risk of pancreatic cancer.” Journal of the National Cancer Institute 101.6 (2009): 424-431. ABO Blood Group and the Risk of Pancreatic Cancer

11. Amundadottir, Laufey, et al. “Genome-wide association study identifies variants in the ABO locus associated with susceptibility to pancreatic cancer.” Nature genetics 41.9 (2009): 986-990.…

12. Wolpin, Brian M., et al. “Pancreatic cancer risk and ABO blood group alleles: results from the pancreatic cancer cohort consortium.” Cancer research 70.3 (2010): 1015-1023. http://cancerres.aacrjournals.or…

13. Wolpin, Brian M., et al. “Variant ABO blood group alleles, secretor status, and risk of pancreatic cancer: results from the pancreatic cancer cohort consortium.” Cancer Epidemiology Biomarkers & Prevention 19.12 (2010): 3140-3149.…

14. Rizzato, Cosmeri, et al. “ABO blood groups and pancreatic cancer risk and survival: results from the PANcreatic Disease ReseArch (PANDoRA) consortium.” Oncology reports 29.4 (2013): 1637-1644. ABO blood groups and pancreatic cancer risk and survival: Results from the PANcreatic Disease ReseArch (PANDoRA) consortium

15. Aird, Ian, H. He Bentall, and JA Fraser Roberts. “Relationship between cancer of stomach and the ABO blood groups.” British Medical Journal 1.4814 (1953): 799.…

16. Edgren, Gustaf, et al. “Risk of gastric cancer and peptic ulcers in relation to ABO blood type: a cohort study.” American journal of epidemiology 172.11 (2010): 1280-1285. Risk of Gastric Cancer and Peptic Ulcers in Relation to ABO Blood Type: A Cohort Study

17. Etemadi, Arash, et al. “Mortality and cancer in relation to ABO blood group phenotypes in the Golestan Cohort Study.” BMC medicine 13.1 (2015): 1. Mortality and cancer in relation to ABO blood group phenotypes in the Golestan Cohort Study

18. Dentali, Francesco, et al. “Non-O blood type is the commonest genetic risk factor for VTE: results from a meta-analysis of the literature.” Seminars in thrombosis and hemostasis. Vol. 38. No. 05. Thieme Medical Publishers, 2012.

19. Dentali, Francesco, et al. “ABO blood group and vascular disease: an update.” Seminars in thrombosis and hemostasis. Vol. 40. No. 01. Thieme Medical Publishers, 2014.

20. Dentali, Francesco, et al. “Relationship between ABO blood group and hemorrhage: a systematic literature review and meta-analysis.” Seminars in thrombosis and hemostasis. Vol. 39. No. 01. Thieme Medical Publishers, 2013.

21. Ohira, T., et al. “ABO blood group, other risk factors and incidence of venous thromboembolism: the Longitudinal Investigation of Thromboembolism Etiology (LITE).” Journal of Thrombosis and Haemostasis 5.7 (2007): 1455-1461.…

22. Franchini, Massimo, and Mike Makris. “Non-O blood group: an important genetic risk factor for venous thromboembolism.” Blood Transfus 11.2 (2013): 164-5.…

23. Spiezia, Luca, et al. “ABO blood groups and the risk of venous thrombosis in patients with inherited thrombophilia.” Blood Transfus 11.2 (2013): 250-253.

24. Franchini, Massimo, and Carlo Bonfanti. “Evolutionary aspects of ABO blood group in humans.” Clinica Chimica Acta 444 (2015): 66-71.

25. Dickey, W., et al. “Secretor status and Helicobacter pylori infection are independent risk factors for gastroduodenal disease.” Gut 34.3 (1993): 351-353. Secretor status and Helicobacter pylori infection are independent risk factors for gastroduodenal disease.

26. Tirumalai Kamala’s answer to Why do my American friends get sick by norovirus every Thanksgiving, but I’ve never seen a Russian citizen gotten sick by norovirus in her homeland?

27. Cserti, Christine M., and Walter H. Dzik. “The ABO blood group system and Plasmodium falciparum malaria.” Blood 110.7 (2007): 2250-2258.…

28. Cserti‐Gazdewich, C. M., W. R. Mayr, and W. H. Dzik. “Plasmodium falciparum malaria and the immunogenetics of ABO, HLA, and CD36 (platelet glycoprotein IV).” Vox sanguinis 100.1 (2011): 99-111.

29. Anstee, David J. “The relationship between blood groups and disease.” Blood 115.23 (2010): 4635-4643.

30. Rowe, J. Alexandra, et al. “Blood group O protects against severe Plasmodium falciparum malaria through the mechanism of reduced rosetting.” Proceedings of the National Academy of Sciences 104.44 (2007): 17471-17476.…

31. Fry, Andrew E., et al. “Common variation in the ABO glycosyltransferase is associated with susceptibility to severe Plasmodium falciparum malaria.” Human molecular genetics 17.4 (2008): 567-576. Common variation in the ABO glycosyltransferase is associated with susceptibility to severe Plasmodium falciparum malaria

32. Timmann, Christian, et al. “Genome-wide association study indicates two novel resistance loci for severe malaria.” Nature 489.7416 (2012): 443-446.

What are the illnesses that can be reliably detected by odor?


, , ,

Just analyze someone’s breath, mucus, saliva, sweat or urine and diagnose whether they have diabetes, cancer, COPD, IBD, Clostridium difficile infection (CDI), Tuberculosis (TB), or any other disease for that matter. Now that would be a seismic revolution in medicine. For one, fully non-invasive in complete contrast to the present-day staples, needles and blood draws, and the pain and fear they entail. Could also be done as often as possible, even when asleep or anesthetized during surgery, even in real-time, as point-of-care, i.e., truly portable and thus truly mobile. Underlying idea is the body’s physiological emanations reliably communicate unique signatures of underlying diseases in the form of singular mixes of volatile organic compounds (VOCs), i.e., the human ‘volatilome‘. The reality, OTOH, is a sharp, painful thud since the ground reality is one where most of these possibilities remain not even remotely feasible in the near future.

Volatilome historians point to the French chemist, Antoine Lavoisier, as the modern inspiration for diagnosis using exhaled breath (1). He showed the body produces and exhales carbon dioxide. In turn this became the basis for Capnography, monitoring the concentration pressure or partial pressure of carbon dioxide in respiratory gases, the most common breath test. Antiquated roots notwithstanding, using unique signatures from breath and other emanations for disease diagnosis still remains deep in research mode and far from much practical utility. A 2014 review lists a total of only 7 US FDA-approved breath-related tests (see below from 2).

Obstacles To Widespread Non-invasive Sampling of Body Emanations For Disease Diagnosis

I. Unlike Animal Olfaction, Human Technology’s Remained Too Constrained In Choice Of What To Analyze

Default, standard analytical approach to sampling and analyzing compounds present in emanations is to rely on ‘headspace’ analysis (3, see below from 4), jargon that means sampling what is already in the gas phase, i.e., already volatile in the material, rather than attempt to extract compounds of interest from it.

Problem is this can cut off too thin a slice of the pie. This shows up in the results since technical approaches continue to fail to mimic what animals do so effortlessly when they use smell to communicate, forage and assess the health of those around them. This brings us to animals and their remarkable capacity to sniff out disease so much so that anecdotal reports suggest they can even be far more accurate in diagnosing human diseases compared to human technologies.

Two diseases with substantial research on animals successfully sniffing them out in humans are Skin Cancers by dogs and TB by Giant African pouched rats.

IA. Dogs Can Spontaneously Detect Human Skin Cancer & Can Also Be Trained To Detect Clostridium difficile infection (CDI)

In 1989, the Lancet published what is perhaps the first modern report of a dog sniffing out its owner’s melanoma (See below from 5)

In 2001, a follow-up anecdotal study (6) added two other case reports of dogs accurately sniffing out skin cancer lesions,

  • One, a London, UK report on Parker, a pet labrador who sniffed it out on a 66 year old man’s left thigh. The patient had developed an eczema patch there. Treated unsuccessfully by topical steroids and antifungals, it grew slowly over 18 years. In 1994, Parker became a family member. Around 1999, Parker started to persistently push his nose against the patient’s trouser leg and sniff the lesion beneath it, i.e., could smell something about the lesion even through clothing. This induced the patient to re-consult his family physician. The lesion was excised in September 2000 and histology showed it to be a fully excised basal cell carcinoma. Once lesion was fully removed, Parker no longer showed interest in that area.
  • The other, by George, a Florida, USA, K-9 unit schnauzer trained by his retired handler to recognize in vitro malignant melanoma samples. A local dermatologist had read the original 1989 Lancet case report and teamed up with the handler to see if such a result was repeatable with another dog. When George was introduced to a patient with several moles considered cancer-free, he went ‘crazy’ over one particular mole, which when excised confirmed ‘early malignant disease’.

Authors of this second case report speculated dogs might also be able to detect odors associated with specific diseases such as TB and Ebola.

In 2012, The BMJ published a proof of principle study on Cliff, a 2 year old beagle trained to sniff out human Clostridium difficile infection (CDI) remarkably accurately (7).

Such reports have provoked more systematic studies, which conclude that dogs could detect unique odors emanating from human melanoma and other cancers (8, 9, 10).

Problem is clinical scope for using dogs to diagnose diseases is limited given the costs, effort, space and time required to train sniffer dogs to detect various diseases (11).

IB. Giant African Pouched Rats Can Be Trained To Reliably Detect Human Tuberculosis (TB)

Used to diagnose lung TB in resource-poor settings, the antiquated microscopic Ziehl–Neelsen stain is a standard method for detecting Mycobacterium tuberculosis in Sputum (coughed up mucus). A few studies suggest trained Giant pouched rat are not just as sensitive and accurate but also able to process >50 times more samples per day compared to a lab technician, i.e., much more economical (12). The WHO recommends microscopists not analyze more than an average of 20 samples per day to minimize misdiagnoses (13) while two trained giant African pouched rats could reach a total of 70 consensus results in 32 minutes over 2 sessions each. This means trained rats could screen larger populations in a much shorter time meaning faster TB diagnoses and hence potential for reduced TB transmission, i.e., a potentially enormous public health benefit.

The video below shows how the Belgian social enterprise APOPO trains these rats in Tanzania to accurately diagnose TB from human samples.

II. Inadequate Research Efforts To Deconstruct The Human Volatilome

Clearly animals are able to smell broader, more complex mix of volatile chemicals that the most sophisticated chemical extraction techniques used in volatilome analysis miss (3). In order for technologies to be able to replicate what animals seem to do so effortlessly, research needs to systematically unravel the human volatilome and establish a reference base of what that looks like in health in order to be able to discern and diagnose cause of ill-health simply from analyzing someone’s emanations.

What compounds are present in normal breath, urine, skin emanations, saliva, blood and feces? A compendium of the healthy human volatilome was first described only in 2014 (14), meaning a foundational study has come along only in the very recent past. This study is foundational for the following reasons,

  • It identified compounds in breath (874), urine (279), skin emanations (504), saliva (353), blood (130), feces (381).
  • It classified these compounds by their CAS Registry Number (CAS), unique numerical identifier assigned to every chemical substance described in the published scientific literature.
    • Hundreds of peer-reviewed scientific papers are routinely published on the human volatilome. Problem is there is as yet no standardization of procedures or data reporting. As a result, the literature is awash in duplicates.
    • Does exhaled breath really have ~3000 different compounds? Umm, looks like it’s less than a fourth of that.
  • This 2014 paper (14) is thus the first step in the right direction, namely to consolidate, synthesize and whittle down published information into a potentially universal ‘megatable’ of compounds present in healthy human emanations.

III. Too Much Technical Sensitivity Can Sometimes Be Too Much Of A Good Thing

  • Iterations over decades have vastly improved sensitivity of state-of-the-art volatilome analysis methods like Gas chromatography–mass spectrometry (GC-MS) and Proton-transfer-reaction mass spectrometry (PTR-TOF-MS), Selected-ion flow-tube mass spectrometry (SIFT-MS) and other techniques such that they can easily measure ~1000 compounds.
    • Problem is most volatilome studies investigate a handful of subjects, not the thousands necessary to validate variables that are different between those with or without diseases.
    • One review (1) suggests number of subjects should be >5X the number of analytes measured, clearly something that adds prohibitive cost to such studies but not doing so increases chances for what they call ‘voodoo correlations’, an issue compounded by dividing test populations further into sub-groups.
  • Volatilome data are also not standardized, neither are the procedures (1, 3, 14, 15, 16, 17, 18). This makes meta-analyses, i.e., comparison of data across multiple studies, well-nigh impossible.
  • As with so many topics in biomedical research, human volatilome studies have heretofore paid scant attention to human Microbiota (19), how it shapes the human volatilome and how that process not only differs between health and disease but also yields different outcomes, i.e., different volatile signatures (1, 3).

That said, there are several diseases with candidate volatile biomarkers that await final validation (see tables below from 20). The sky’s very much the limit for diagnosing diseases through their distinctive odors.


1. Amann, Anton, et al. “The human volatilome: volatile organic compounds (VOCs) in exhaled breath, skin emanations, urine, feces and saliva.” Journal of breath research 8.3 (2014): 034001.

2. Amann, Anton, et al. “Analysis of exhaled breath for disease detection.” Annual Review of Analytical Chemistry 7 (2014): 455-482.

3. Kimball, Bruce A. “Volatile metabolome: problems and prospects.” (2016).…

4. Restek, A. “Technical Guide for Static Headspace Analysis Using GC.” Restek Corp (2000): 11-12.…

5. Williams, Hywel, and Andres Pembroke. “Sniffer dogs in the melanoma clinic?.” The Lancet 333.8640 (1989): 734.

6. Church, John, and Hywel Williams. “Another sniffer dog for the clinic?.” The Lancet 358.9285 (2001): 930.…

7. Bomers, Marije K., et al. “Using a dog’s superior olfactory sensitivity to identify Clostridium difficile in stools and patients: proof of principle study.” (2012): e7396.…

8. Pickel, Duane, et al. “Evidence for canine olfactory detection of melanoma.” Applied Animal Behaviour Science 89.1 (2004): 107-116.…

9. Moser, Emily, and Michael McCulloch. “Canine scent detection of human cancers: a review of methods and accuracy.” Journal of Veterinary Behavior: Clinical Applications and Research 5.3 (2010): 145-152.…

10. Jezierski, Tadeusz, et al. “Study of the art: canine olfaction used for cancer detection on the basis of breath odour. Perspectives and limitations.” Journal of breath research 9.2 (2015): 027001.…

11. Buljubasic, Fanis, and Gerhard Buchbauer. “The scent of human diseases: a review on specific volatile organic compounds as diagnostic biomarkers.” Flavour and Fragrance Journal 30.1 (2015): 5-25.

12. Mgode, Georgies F., et al. “Diagnosis of tuberculosis by trained African giant pouched rats and confounding impact of pathogens and microflora of the respiratory tract.” Journal of clinical microbiology 50.2 (2012): 274-280. Diagnosis of Tuberculosis by Trained African Giant Pouched Rats and Confounding Impact of Pathogens and Microflora of the Respiratory Tract

13. World Health Organization, et al. “Management of tuberculosis: training for district TB coordinators.” (2005).…

14. de Lacy Costello, Ben, et al. “A review of the volatiles from the healthy human body.” Journal of breath research 8.1 (2014): 014001.…

15. Pereira, Jorge, et al. “Breath analysis as a potential and non-invasive frontier in disease diagnosis: an overview.” Metabolites 5.1 (2015): 3-55. Breath Analysis as a Potential and Non-Invasive Frontier in Disease Diagnosis: An Overview

16. Boots, Agnes W., et al. “Exhaled molecular fingerprinting in diagnosis and monitoring: validating volatile promises.” Trends in molecular medicine 21.10 (2015): 633-644.…

17. Bikov, Andras, Zsófia Lázár, and Ildiko Horvath. “Established methodological issues in electronic nose research: how far are we from using these instruments in clinical settings of breath analysis?.” Journal of breath research 9.3 (2015): 034001.

18. Scarlata, Simone, et al. “Exhaled breath analysis by electronic nose in respiratory diseases.” Expert review of molecular diagnostics 15.7 (2015): 933-956.

19. Dietert, Rodney Reynolds, and Ellen Kovner Silbergeld. “Biomarkers for the 21st century: listening to the microbiome.” Toxicological Sciences (2015): kfv013. Listening to the Microbiome

20. Kataoka, Hiroyuki, et al. “Noninvasive analysis of volatile biomarkers in human emanations for health and early disease diagnosis.” Bioanalysis 5.11 (2013): 1443-1459.…

Thanks for the R2A, Jonathan Brill.

What are the key immunological markers of successful cancer immunotherapy?



The two sides to successful cancer immunotherapy are

  • Unique features of the cancer that engender artificially driving cancer-specific immunity (cytotoxic CD8 T cells for example) and/or offer paths exploitable by immunotherapy (antibodies against checkpoint inhibitors such as CTLA-4 or PD-1 for example).
  • Potential for effective cancer-specific immune responses.

Cancer Features That Favor Immunotherapy Success

Research shows certain cancer features are conducive to helping initiate and/or sustain effective anti-cancer immunity.

These include

  • High level of genomic instability and mutational load. Data supporting this feature observed in immunotherapy success cases against non-small cell lung cancer (1), melanoma (2), metastatic melanoma (3, 4) and colorectal cancer (5).
    • To ensure the immune system doesn’t attack the body itself, most tissue antigen-specific T (and B) cells never make it through developmental bottlenecks. In other words, normal immunological tolerance processes that safeguard body from attack by its own immune system are also a natural barrier to strong anti-tumor immunity.
    • What is a cancer cell? A cell that some time in the past went rogue, no longer subject to growth control. What does that signify to the immune system though? Does it remain a cell the immune system’s been programmed to tolerate or something different that it can recognize and respond to? Having started out as a normal tissue cell, most of what a tumor cell expresses is the same as that tissue.
    • The crux that determines whether or not immunotherapy can even be harnessed to rid of a tumor is how antigenically different it’s become from a normal cell.
    • More antigenically similar a tumor to normal tissue, lower the chances of tumor-specific immune cells. After all, to even be able to make effective anti-tumor immune responses in the first place, one needs tumor-specific T cell (CD4 and CD8).
    • Since the human body’s immune system is geared to push through T and B cells specific for antigens not expressed by it in the normal course of a lifetime, neoantigen-specific T and B cells are to be expected.
    • Thus, more neoantigens a tumor expresses (6), greater the likelihood they can be ‘seen’ by the immune system (7, 8), greater the likelihood of higher frequency of neoantigen-specific cytotoxic CD8 T cells, and greater the chances of being able to artificially stoke effective tumor-specific immunity.
  • Presence of Tumor-infiltrating lymphocytes (TIL). More TILs, especially cytotoxic CD8s, more immunogenic the tumor and tumor-associated cells, i.e., capable of eliciting anti-tumor-specific immunity (see below from 9). Most cancer immunotherapies try to maximize targeted killing by cytotoxic CD8 T cells, which is considered the main, most effective anti-tumor immune cell type and response. Such approaches work poorly in non-immune cell solid tumor patients who have few such cells to start with, for e.g., prostate cancer.
  • OTOH, tumors with few or no TILs could mean (see below from 10)
    • Tumor actively restricts their entry. For e.g., by releasing specific chemokines such as nitrated CCL2.
    • Tumor vasculature expresses specific molecules such as Fas ligand capable of causing T cell apoptosis (death) or Prostaglandin E2 capable of blocking their effector response.
    • Tumor may maintain low oxygen levels in and around it, Hypoxia (medical). In turn, such conditions promote Programmed cell death protein 1 expression on tumor-associated antigen-presenting cells. PD-1 binding to PD-L1 on CD8 T cells inhibits their response.
    • Tumor microenvironment may favor accumulation of metabolites such as Indoleamine 2,3-dioxygenase (IDO) which inhibit CD8 T cell response.
    • Tumor may specifically recruit certain fibroblasts and B cells to its vicinity to inhibit CD8 T cell entry and responses, respectively.
  • In other words, successful cancer immunotherapy against solid tumors has to be carefully engineered in order to prevail over numerous natural physical and functional obstacles.

Features of effective anti-cancer immunity

  • More TILs and more CD8s among such TILs.
  • High level of PD-L1 expression. Even though high PD-L1 expression in tumors inhibits effective anti-tumor immunity by binding to PD-1 on CD8 T cells and inhibiting them, targeting PD-1 or PD-L1 through specific mAbs (Monoclonal antibody) is akin to lifting the brakes from these cells, unleashing effective anti-tumor immunity (11, 12).
  • Interferon gamma is an important cytokine in the cytotoxic T cell’s armamentarium. A 2016 study (13) finds melanomas resistant to Ipilimumab, mAb against checkpoint inhibitor CTLA-4, lose IFN-gamma signaling capacity, i.e., plausible reason for their resistance.

Assays And Biomarkers Used To Measure Anti-Tumor Immunity During Immunotherapy

Cellular Techniques To Assess Immune Status Within Tumors

  • IHC (immunohistochemistry) is an old workhorse that’s used to count the number of TILs in a tumor tissue using anti-CD8 antibodies. Now anti-PD-L1 antibodies are also being used to assess PD-L1 expression as a biomarker (14) for how useful anti-PD-1 or -PD-L1 antibodies might be in releasing the brakes to unleash anti-tumor immune responses from TILs present within it. Such assessments have caveats because different studies showed different predictive power of anti-PD-L1 IHC (15, 16).
    • Different studies used different antibodies and used different thresholds for assessing positivity (17).
    • An inducible cell-surface molecule, PD-L1’s presence or absence in archival tissue samples can’t fully predict its real-time status.
  • Immunoscore (18), a relatively new pathology algorithm developed by French pathologist, Jérôme Galon, uses digital pathology to minimize variability and provides quantitative data on T cells within a tumor, not just in its center but also in its margins, something that improves prognostic accuracy. This approach is slowly working its way through to widespread validation and maybe eventual acceptance (19).
  • Currently hobbled by technical limitations and interpretation difficulties (20), multiplex IHC, i.e., trying to assess multiple tissue- and cell-specific markers simultaneously, would provide more information not only on numbers of immune cells within tumors but also their spatial organization. Sequential multiplex IHC is an approach to make this technology workable (21) but it’s still far from practical utility.
  • Cheap, long-lasting, FFPE (formalin fixed paraffin embedded) tissue sections are a mainstay in pathological diagnoses. Mass spectrometry based FFPE multiplexing is an attempt to marry this ancient technique with a more modern molecular analytical tool to exponentially increase the number of markers assessed to as many as 100 (22). A 2014 study (23) used this approach to simultaneously image as many as 32 breast cancer-associated hormonal and immunological proteins.
  • Flow cytometry & Mass cytometry (CyTOF). Of more value in blood cancers rather than solid tumors, using antibodies labeled with rare metals, CyTOF (cytometry in time-of-flight) combines flow cytomtery and mass spectromtery. A 2015 study simultaneously assessed 15 surface and 16 intracellular proteins in leukemia and showed mismatch between surface and intracellular states of some of these proteins (24) while also being able to identify that cell cycle differences between leukemia stem cells could influence their response to therapy (25).

Genomic Techniques To Assess Immune Status Within Tumors

  • Exome sequencing (Whole Exome Sequencing, WES) of tumor tissue allows its mutational load estimation (1, 2, 3). Combining this with algorithms that help predict likelihood that peptide sequences bind to HLA (Human leukocyte antigen) or help predict likelihood TCRs (T cell receptors) will recognize them could improve Rx outcomes in a more individualized manner.
  • DNA mismatch repair (MMR) deficiency could be a biomarker of response to PD-1 based immunotherapy (26), especially in colorectal cancers (5).
  • T cell receptor (TCR) sequencing can monitor changes in T cell populations within a tumor over the course of Rx. Differences between responders and non-responders might identify those more likely to benefit from Rx (27).
  • Describing a tumor’s transcriptional activity, RNA sequencing is to RNA what WES is to DNA (28).
  • While all these approaches have been attempted on tumor tissues, newer approaches seek to apply them at the single cell level. One such approach showed for example that mutational patterns in each cancer cell in acute myeloid leukemia patients was different (29).
  • Results such as these reveal why even the latest immunotherapies end up benefiting only a handful of patients and why realistic cancer cure may only come from individualized, and hence extremely costly, efforts.
  • Most of these approaches are invasive, requiring access to tumor itself. A non-invasive counterpart tries to take advantage of tumor metabolite spillover into bloodstream.
    • Exosome (vesicle) are 50 to 100 nm membrane-bound vesicles secreted by many cells, including tumor and immune cells. They’re thought to be a way for cells to communicate over short and long distances by exchanging genetic and protein material. An as-yet unpublished 2015 study that used analysis of circulating exosomes to assess responses to cancer immunotherapy found differences between responders and non-responders (30), suggesting such differences could be used as predictive biomarkers, if found reproducible.

See composite pictorials below from 9, 28 on cellular and genomic approaches currently used or proposed to be used to monitor anti-tumor immunity during immunotherapy in some clinical trials.

Practical Obstacles to Studying Human Anti-Tumor Immunity

  • Kind of tumor tissue available for assessment makes a huge difference, specifically whether tissue is archival or fresh (see below from 9). Advantages of the former are to the patient, making another invasive biopsy unnecessary. OTOH, disadvantages include changes to cell, protein and gene expression profiles depending on how archival tissue’s been preserved. Obviously, fresh tumor tissue reflects current disease state more accurately, particularly changes in response to Rx.
  • Since tumor and immune cell features are dynamic, not static, sampling over time may yield more accurate information. Obviously this increases cost and may also increase risk to patient.
  • PD-L1 represents an excellent example of pitfalls inherent to choice of tumor tissue and sampling frequency used in cancer immunotherapy decision making. On the one hand, it’s now well appreciated that higher the PD-L1 expression in tumors, better the chances of Cancer immunotherapy. However, most of the time, PD-L1 expression is assessed on archival tissue. Since PD-L1 is an inducible molecule, making Rx decisions on results from archival tissue thus carries a two-fold burden, one, passage of time since that sample was taken, and two, dynamic changes in PD-L1 expression in tumor.
  • Amount of tissue available is another important variable (see below from 9).
    • Core biopsies are too small to enable meaningful immunological assessments such as topographic TIL enumeration, i.e., how many are present in the tumor center versus its invasive margins.
    • Tumors can also be highly heterogeneous, not just at different sites (31) but also within the same tumor (29, 32).
    • More sampling at more places gives a more accurate picture but again at higher cost and risk to the patient.


1. Rizvi, Naiyer A., et al. “Mutational landscape determines sensitivity to PD-1 blockade in non–small cell lung cancer.” Science 348.6230 (2015): 124-128.

2. Snyder, Alexandra, et al. “Genetic basis for clinical response to CTLA-4 blockade in melanoma.” New England Journal of Medicine 371.23 (2014): 2189-2199.…

3. Van Allen, Eliezer M., et al. “Genomic correlates of response to CTLA-4 blockade in metastatic melanoma.” Science 350.6257 (2015): 207-211.

4. Hugo, Willy, et al. “Genomic and transcriptomic features of response to anti-PD-1 therapy in metastatic melanoma.” Cell 165.1 (2016): 35-44.…

5. Le, Dung T., et al. “PD-1 blockade in tumors with mismatch-repair deficiency.” New England Journal of Medicine 372.26 (2015): 2509-2520.…

6. Gubin, Matthew M., et al. “Tumor neoantigens: building a framework for personalized cancer immunotherapy.” The Journal of clinical investigation 125.9 (2015): 3413-3421. Tumor neoantigens: building a framework for personalized cancer immunotherapy

7. Schumacher, Ton N., and Robert D. Schreiber. “Neoantigens in cancer immunotherapy.” Science 348.6230 (2015): 69-74.…

8. Schumacher, Ton N., and Nir Hacohen. “Neoantigens encoded in the cancer genome.” Current Opinion in Immunology 41 (2016): 98-103.

9. Wargo, Jennifer A., et al. “Monitoring immune responses in the tumor microenvironment.” Current opinion in immunology 41 (2016): 23-31.…

10. Joyce, Johanna A., and Douglas T. Fearon. “T cell exclusion, immune privilege, and the tumor microenvironment.” Science 348.6230 (2015): 74-80.

11. Tumeh, Paul C., et al. “PD-1 blockade induces responses by inhibiting adaptive immune resistance.” Nature 515.7528 (2014): 568-571.…

12. Ribas, Antoni, et al. “PD-1 blockade expands intratumoral memory T cells.” Cancer immunology research 4.3 (2016): 194-203. http://cancerimmunolres.aacrjour…

13. Gao, J. et al. Loss of IFN-g pathway genes in tumor cells as a mechanism of resistance to anti-CTLA-4 therapy. Cell, 2016.

14. Fusi, Alberto, et al. “PD-L1 expression as a potential predictive biomarker.” The Lancet Oncology 16.13 (2015): 1285-1287.

15. Herbst, Roy S., et al. “Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients.” Nature 515.7528 (2014): 563-567. http://www.livewell-bioscience.c…

16. Carbognin, Luisa, et al. “Differential activity of nivolumab, pembrolizumab and MPDL3280A according to the tumor expression of programmed death-ligand-1 (PD-L1): sensitivity analysis of trials in melanoma, lung and genitourinary cancers.” PloS one 10.6 (2015): e0130142.…

17. Gadiot, Jules, et al. “Overall survival and PD‐L1 expression in metastasized malignant melanoma.” Cancer 117.10 (2011): 2192-2201. Overall survival and PD-L1 expression in metastasized malignant melanoma

18. Galon, Jérôme, et al. “Type, density, and location of immune cells within human colorectal tumors predict clinical outcome.” Science 313.5795 (2006): 1960-1964.…

19. Galon, Jérôme, et al. “Towards the introduction of the ‘Immunoscore’in the classification of malignant tumours.” The Journal of pathology 232.2 (2014): 199-209.…

20. Stack, Edward C., et al. “Multiplexed immunohistochemistry, imaging, and quantitation: a review, with an assessment of Tyramide signal amplification, multispectral imaging and multiplex analysis.” Methods 70.1 (2014): 46-58.…

21. Tsujikawa, Takahiro, et al. “Multiplex immunohistochemistry for immune profiling of HPV-associated head and neck cancer.” Journal for immunotherapy of cancer 3.2 (2015): 1. Multiplex immunohistochemistry for immune profiling of HPV-associated head and neck cancer

22. Giesen, Charlotte, et al. “Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.” Nature methods 11.4 (2014): 417-422.…

23. Angelo, Michael, et al. “Multiplexed ion beam imaging of human breast tumors.” Nature medicine 20.4 (2014): 436-442.…

24. Levine, Jacob H., et al. “Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis.” Cell 162.1 (2015): 184-197.…

25. Behbehani, Gregory K., et al. “Mass cytometric functional profiling of acute myeloid leukemia defines cell-cycle and immunophenotypic properties that correlate with known responses to therapy.” Cancer discovery 5.9 (2015): 988-1003. http://cancerdiscovery.aacrjourn…

26. Kelderman, Sander, Ton N. Schumacher, and Pia Kvistborg. “Mismatch repair-deficient cancers are targets for anti-PD-1 therapy.” Cancer cell 28.1 (2015): 11-13.

27. Postow, Michael A., et al. “Peripheral T cell receptor diversity is associated with clinical outcomes following ipilimumab treatment in metastatic melanoma.” Journal for immunotherapy of cancer 3.1 (2015): 1. Peripheral T cell receptor diversity is associated with clinical outcomes following ipilimumab treatment in metastatic melanoma

28. Dijkstra, Krijn K., et al. “Genomics-and Transcriptomics-Based Patient Selection for Cancer Treatment With Immune Checkpoint Inhibitors: A Review.” JAMA oncology (2016).

29. Paguirigan, Amy L., et al. “Single-cell genotyping demonstrates complex clonal diversity in acute myeloid leukemia.” Science translational medicine 7.281 (2015): 281re2-281re2.…

30. Hurley, J., et al. “452 Profiling exosomal mRNAs in patients undergoing immunotherapy for malignant melanoma.” European Journal of Cancer 51 (2015): S96.

31. Reuben, Alexandre, et al. “Molecular and immune heterogeneity in synchronous melanoma metastases.” Journal for immunotherapy of cancer 3.2 (2015): 1. Molecular and immune heterogeneity in synchronous melanoma metastases

32. Zhang, Jianjun, et al. “Intratumor heterogeneity in localized lung adenocarcinomas delineated by multiregion sequencing.” Science 346.6206 (2014): 256-259.…

What are some ways to test if you have good gut bacteria?

What constitutes good gut bacteria? What’s their benchmark? We have no clue. Economics of gene sequencing technology means fecal microbiome sequencing costs (1) a fraction of what it did just a few years back. Predictably, companies offering to sequence them have mushroomed, for a price of course. Anyone with a handful of disposable US dollars can get their poop bacteria sequenced but what do those results even mean?

  • Should poop contain ~55% Firmicutes, apparently the same as Michael Pollan or 74%, same as the author of this Newsweek piece (2)?
  • What’s the value of a one-time poop bacteria sequencing? Isn’t that just a snapshot?
  • Studies show poop bacterial composition changes rapidly not just with diet (3) but also seasonally (4) so what’s the predictive value of such a snapshot anyway?
  • What does poop bacteria even represent?
    • Isn’t poop bacteria sequenced so much just because it’s easier to access?
    • Doesn’t it really only represent distal colon bacteria supported by current diet?
    • What about what’s in the stomach, duodenum, jejunum, ileum, and proximal and transverse colon, and how they relate to gut and overall health? Don’t we need invasive biopsies to accurately access bacteria in other GI tract compartments?
    • What can we extrapolate from what’s in poop to what should be in other parts of the GI tract? Anything? Nothing?

All this to say a lot of data on poop microbiome’s being generated simply because it can be, not because anyone has a clue what any of it means nor a clue what constitutes good gut bacteria.

To top this litany of shortcomings and dubious value of current attempts to benchmark gut bacteria using fecal microbiome sequencing, at least one randomized placebo-controlled study (5, 6) not only reveals novel, incalculable curative powers of Placebo but also casts doubt on currently accepted notions of ‘good’ and ‘bad’ gut bacteria.

  • A study across two US academic medical centers, Montefiore Medical Center in the Bronx, New York and the Miriam Hospital, Providence, Rhode Island, both well-known for their expertise in Fecal microbiota transplant (FMT) (5, 6).
  • 46 patients with recurrent Clostridium difficile infection (CDI) were randomly assigned to receive either donor or autologous (their own) poop microbiota, i.e., Placebo.
  • 91% (20/22) of those who got donor poop were durably cured based on a standard definition. Expected so no surprise.
  • The absolute shocker? 63% (15/24) who got their own poop microbiota transplanted back also had durable cure. Rub eyes and read again. What? Patients with a serious GI tract infection were given back their own presumably disease-associated gut bacteria and they got cured?
  • Though there were striking inter-center differences in this Placebo effect, 9/10 (90%) placebo cases in the New York center cured versus only 6/14 (~43%) in the Rhode Island center, and perhaps associated patient population differences between these two centers, those aren’t germane to the central issue, namely, a GI tract disease cured from simply taking out and putting fecal bacteria back into C.diff patients.

Thus, even so-called ‘bad’ gut bacteria turn out to be not so cut and dry, a result that only underlines how little we currently know about gut bacteria, good, bad or anything in between. Best one could then say is absence of persistent and serious health problems, especially gastrointestinal, is evidence of having good gut bacteria. Absence of skin problems, no autoimmunities or mental health issues would be icing on the cake.


1. 16S rRNA sequencing

2. Newsweek, Roxane Khamsi, July 17, 2014. Gut Check

3. David, Lawrence A., et al. “Diet rapidly and reproducibly alters the human gut microbiome.” Nature 505.7484 (2014): 559-563. Diet rapidly and reproducibly alters the human gut

4. Davenport, Emily R., et al. “Seasonal variation in human gut microbiome composition.” PloS one 9.3 (2014): e90731.…

5. Kelly, Colleen R., et al. “Effect of Fecal Microbiota Transplantation on Recurrence in Multiply Recurrent Clostridium difficile Infection: A Randomized Trial.” Annals of Internal Medicine (2016).

6. Fecal Transplant for Relapsing C. Difficile Infection

What is the function of bacteria in the human mouth?


, , , , , ,

Most of the data available so far identifies bacterial species that tend to be associated with healthy versus diseased oral cavities but not much is known about exactly what health-associated ones do apart from keeping out the disease-associated ones.

Oral Cavity, A Complex Ecosystem Of Several Specialized Ecological Niches

Gaining insight begins with the appreciation that the oral cavity is a complex habitat further sub-divided into distinct smaller ones ranging from the non-keratinized Oral mucosa to the keratinized Tongue and gingiva, i.e., Gums, as well as Tooth enamel and a variety of dental implants. Thus, depending on its proximity to the gum line, dental plaque is either supra- or subgingival, and the bacteria that inhabit these two regions are different, supragingival plaque dominated by gram-positive Streptococci while subgingival by gram-negative anaerobic bacteria (1).

Since the oral cavity is exposed to the outside world, these surfaces are colonized soon after birth, with some evidence suggesting vertical (mother-to-baby) transmission (2) as well as similarities between family members (3). Stable inhabitants, formally called autochthonous, establish Biofilm, a kind of super-organism consisting of cooperating microbes. Being open, oral cavity also gets plenty of visitors, transients, formally called allochthonous.

Older studies suggested oral cavity diseases are associated with changes in microbial diversity (1).

Gum diseases, i.e., Periodontal pathology, and tooth decay, i.e., Dental caries, are associated with increase (4, 5, 6) and decrease (7), respectively, in microbial diversity. More recent state-of-the-art molecular approaches (8, 9) confirm these decades-old findings. This implies oral cavity disease isn’t so much a matter of presence or absence of certain microbes since those with disease-causing potential, i.e., pathobionts, are present even in health (10) but rather about their relative proportions in complex biofilms.

Colonisation resistance Is A Key Feature Of Healthy Oral Cavity Microbiota

Like other microbe-associated body sites, the oral cavity too is a series of specialized niches occupied by specific microbes capable of specialized functions both necessary and predicated on some inherent properties of these niches. The ones who establish stable presence in the form of complex, multi-species biofilms dominate their specific niches by preventing others, including pathogens, from establishing themselves, i.e., Colonization Resistance (11). When the oral cavity is stably colonized by beneficial microbe biofilms, it’s healthy. Instability in beneficial microbe colonization is a weakness that’s then exploited by more harmful and even pathogenic species to dominate oral cavity biofilms, with the outcome either tooth decay (Dental Caries) or gum disease (Periodontitis).

Tooth decay (Dental caries) is associated with certain species of Streptococci such as Streptococcus mutans and species of Lactobacilli (12) while subgingival anaerobes establish their communities within periodontal pockets, some of whom such as Porphyromonas gingivalis are associated with gum disease (Periodontitis) (13, 14).

Obviously diet profoundly influences not only which bacterial species stably establish within oral biofilms but also dominate.

  • Thus though S. mutans is part of normal oral microbiota (15), it doesn’t dominate in healthy oral cavities.
  • However, its capacity to metabolize sucrose more efficiently compared to other oral bacteria (16) gives it a competitive advantage in the oral cavities of those who predominantly consume the highly processed, sucrose-heavy ‘Western’ diet.
  • S. mutans can also convert sucrose to adherent glucans, which helps it to attach more strongly to teeth (17).
  • S. mutans also rapidly converts sucrose to lactic acid, giving it an added selective advantage owing to its intrinsic capacity to withstand such acidic environments (18, 19).
  • These properties may help S. mutans and Lactobacilli dominate in tooth decay, i.e., dental caries, the latter because they metabolize lactic acid generated by S.mutans.

Similar studies done decades apart, scraping plaques from people with or without gum disease and culturing the bacteria that grew out with bacterial species associated with gum disease showed plaques from people without gum disease can inhibit growth of gum disease-associated bacterial species (20, 21). How?

  • Streptococcus sanguinis is considered an inhabitant of normal dental plaque. Less acid-tolerant than its presumed niche competitor, S. mutans, S. sanguinis produces Hydrogen peroxide, toxic for S. mutans, which typically lacks capacity to effectively neutralize it (22, 23). Thus, dental plaques rich in S. sanguinis contain relatively lower proportions of dental caries-associated S. mutans and periodontitis-associated P. gingivalis (20).
  • Veillonella species (24) and S. oligofermentans (25) readily metabolize lactic acid secreted by S. mutans. A revealing window into how inter-species competition can engender colonization resistance, S. oligofermentans not only utilizes S. mutans-generated lactic acid but converts it into hydrogen peroxide, highly toxic to the latter (26).
  • Streptococcus gordonii offers another plausible example of colonization resistance tactic. Another inhabitant of healthy oral cavities, in vitro it could prevent stable S. mutans colonization by inactivating one of its important resistance mechanisms , ability to synthesize a Quorum sensing molecule, CSP (competence-stimulating peptide) (27). When thus hobbled, S. mutans is much less capable of resisting natural salivary antimicrobial peptides such as Histatin (1).
  • Oral cavity bacteria also secrete Bacteriocin. Proteinaceous toxins, bacteriocins differ from antibiotics, having a much narrower killing spectrum and acting on related organisms (1).

Thus, as long as diet is varied enough to also allow stable plaque colonization by base-producing microbes, acid-producing S. mutans wouldn’t be able to predominate and take over local plaque ecosystem.

Food web Relationships Between Normal Oral Cavity Microbes Help Maintain Their Stability

As is the hallmark of ecosystems consisting of mutually dependent residents, the healthy oral cavity too contains microbes engaged in Food web activities, i.e., metabolic end products of one species used by others.

  • Oral biofilm Streptococci synthesize lactate that Veillonella use (1).
  • S. sanguis and S. oralis are inhabitants of healthy oral biofilms. In vitro culture studies suggest their mutually helpful, i.e., synergistic, capacity to digest mucins helps them more efficiently use such complex host sugars as nutrition (28).
  • Oral cavity is constantly bathed in saliva and gingival crevicular fluid. Composite of products of not just human tissue cells but also microbes, some microbial inhabitants appear to engage in synergistic/mutualistic interactions to overcome inherent handicaps to colonize. This seems to be the case with Actinomyces naeslundii and S. oralis that alone colonize saliva-coated surfaces poorly and yet can form extensive biofilms together by presumably combining their metabolic activities (29).

However, food web processes can also help shift oral biofilms to dominance by more pathogenic species. Though inhabitants of normal oral cavity, P. gingivalis, Fusobacterium nucleatum, Treponema denticola and Tannerella forsythia are also implicated in periodontal disease.

  • In vitro culture studies show P. gingivalis can metabolize succinate produced by T. denticola (30) while the latter can use isobutyric acid secreted by the former (31).
  • Both F. nucleatum and T. forsythia seem to secrete factors that stimulate growth of P. gingivalis (1).

Thus, whether mutualistic interactions of beneficial or harmful bacteria dominate in a given oral cavity is outcome of diet, oral hygiene and host genetic polymorphisms.

Why Knowledge Of Bacteria Function In Healthy Oral Cavity Is Better Gleaned From Older, Not Newer, Studies

Since the 2000s an explosion in molecular biological tools, so-called Omics, has led to a similar explosion in human microbiome studies. Since the oral cavity is one of the most easily accessible of all the GI tract niches, human oral cavity microbiome has become the best characterized in terms of the kinds of bacteria present in healthy versus unhealthy mouths.

  • Since such typically technocratically driven processes focus primarily on generating an avalanche of data and explore no underlying hypotheses, one may wonder whether the healthy human oral cavity microbiome’s function is simply absence of disease. That is to say, given the monumental scale of molecular biology data generated on this topic since at least the mid-2000s, it’s shocking how little is known about what any of it even means.
  • With older prejudices implicitly carried forward, there’s also been no attempt so far to synthesize the roles of bacteria and fungi in healthy oral cavities since fungi like Candida albicans were previously assumed to only represent disease states. Their repeated presence in healthy oral cavities suggests this idea needs revising (32).
  • Even less is known about role of Archaea such as Methanobrevibacter species frequently found in healthy oral cavities. Their increasing identification in gum disease (Periodontal pathology) suggests they too may be involved in such disease processes but how? Only in promoting growth of pathogenic bacterial species (33) or as initiators and perpetuators themselves?
  • Meantime overweening allegiance to novel technologies is powering this entire absurd process forward with the implicit hope that Data mining will uncover hidden patterns allowing certain predictive hypotheses to be made.
  • If past is any predictor of future, the failure of past dependence on novel molecular biological approaches alone to yield predictive insight into complex biological phenomena suggests a similar fate awaits the current giddy immersion in the latest molecular biological toys. A useful and telling example from the recent past is Microarray analysis techniques, the focus of tens of thousands of papers since the late 1990s, which nevertheless yielded little or no improved insight into disease processes nor did they much illuminate possible future predictive approaches to better understand them.
  • Necessity of extrapolating data from in vitro culture studies referenced in this answer is their major caveat. Nevertheless, we’d understand oral cavity-bacteria interactions better with more such experiments, especially in vitro co-cultures of human oral epithelial cells with candidate oral cavity commensals, more so co-cultures with commensal biofilms but such experimental approaches are technically much more challenging compared to powering a few cheek swab or saliva samples through the latest molecular biology apparatus. Hence the current absurd status quo.


1. Kuramitsu, Howard K., et al. “Interspecies interactions within oral microbial communities.” Microbiology and molecular biology reviews 71.4 (2007): 653-670. Interspecies Interactions within Oral Microbial Communities

2. Kobayashi, N., et al. “Colonization pattern of periodontal bacteria in Japanese children and their mothers.” Journal of periodontal research 43.2 (2008): 156-161.…

3. Steenbergen, TJM van, et al. “Intra‐familial transmission and distribution of Prevotella intermedia and Prevotella nigrescens.” Journal of periodontal research 32.4 (1997): 345-350.

4. Löe, Harald, Else Theilade, and S. Börglum Jensen. “Experimental gingivitis in man.” Journal of periodontology 36.3 (1965): 177-187.

5. Listgarten, M. A. “Structure of the Microbial Flora Associated with Periodontal Health and Disease in Man: A Light and Electron Microscopic Study*.” Journal of periodontology 47.1 (1976): 1-18.

6. Syed, S. A., and W. J. Loesche. “Bacteriology of human experimental gingivitis: effect of plaque age.” Infection and immunity 21.3 (1978): 821-829.

7. Simon-Soro, A., et al. “A tissue-dependent hypothesis of dental caries.” Caries research 47.6 (2013): 591-600.

8. Diaz, P. I., A. Hoare, and B. Y. Hong. “Subgingival Microbiome Shifts and Community Dynamics in Periodontal Diseases.” Journal of the California Dental Association 44.7 (2016): 421.…

9. Tanner, A. C., C. A. Kressirer, and L. L. Faller. “Understanding Caries From the Oral Microbiome Perspective.” Journal of the California Dental Association 44.7 (2016): 437.…

10. Jiao, Y., M. Hasegawa, and N. Inohara. “The role of oral pathobionts in dysbiosis during periodontitis development.” Journal of dental research 93.6 (2014): 539-546.…

11. Van der Waaij, D., J. M. Berghuis-de Vries, and J. E. C. Lekkerkerk-Van der Wees. “Colonization resistance of the digestive tract in conventional and antibiotic-treated mice.” Journal of Hygiene 69.03 (1971): 405-411.…

12. Chhour, Kim-Ly, et al. “Molecular analysis of microbial diversity in advanced caries.” Journal of clinical microbiology 43.2 (2005): 843-849. Molecular Analysis of Microbial Diversity in Advanced Caries

13. ÖSterberg, Stic K‐Å., Sara Z. Sudo, and Lars EA Folke. “Microbial succession in subgingival plaque of man.” Journal of periodontal research 11.5 (1976): 243-255.

14. Ximénez‐Fyvie, Laurie Ann, Anne D. Haffajee, and Sigmund S. Socransky. “Microbial composition of supra‐and subgingival plaque in subjects with adult periodontitis.” Journal of clinical periodontology 27.10 (2000): 722-732.…

15. Aas, Jørn A., et al. “Defining the normal bacterial flora of the oral cavity.” Journal of clinical microbiology 43.11 (2005): 5721-5732. Defining the Normal Bacterial Flora of the Oral Cavity

16. Hamada, Shigeyuki, and HUTTON D. Slade. “Biology, immunology, and cariogenicity of Streptococcus mutans.” Microbiological reviews 44.2 (1980): 331.…

17. Gibbons, R. J., and M. Nygaard. “Synthesis of insoluble dextran and its significance in the formation of gelatinous deposits by plaque-forming streptococci.” Archives of oral biology 13.10 (1968): 1249-IN31.

18. Grenier, Daniel. “Antagonistic effect of oral bacteria towards Treponema denticola.” Journal of clinical microbiology 34.5 (1996): 1249-1252. Antagonistic effect of oral bacteria towards Treponema denticola.

19. Doran, A., S. Kneist, and Joanna Verran. “Ecological control: in vitro inhibition of anaerobic bacteria by oral streptococci.” Microbial Ecology in Health and Disease 16.1 (2004): 23-27.…

20. Hillman, J. D., S. S. Socransky, and Myra Shivers. “The relationships between streptococcal species and periodontopathic bacteria in human dental plaque.” Archives of Oral Biology 30.11-12 (1985): 791-795.

21. van Essche, Mark, et al. “Bacterial antagonism against periodontopathogens.” Journal of periodontology 84.6 (2013): 801-811.

22. Carlsson, Jan, and May‐Britt K. Edlund. “Pyruvate oxidase in Streptococcus sanguis under various growth conditions.” Oral microbiology and immunology 2.1 (1987): 10-14.

23. Kreth, Jens, et al. “Competition and coexistence between Streptococcus mutans and Streptococcus sanguinis in the dental biofilm.” Journal of bacteriology 187.21 (2005): 7193-7203. Competition and Coexistence between Streptococcus mutans and Streptococcus sanguinis in the Dental Biofilm

24. Mikx, F. H. M., and J. S. Van der Hoeven. “Symbiosis of Streptococcus mutans and Veillonella alcalescens in mixed continuous cultures.” Archives of Oral Biology 20.7 (1975): 407-410.

25. Tong, Huichun, et al. “Streptococcus oligofermentans inhibits Streptococcus mutans through conversion of lactic acid into inhibitory H2O2: a possible counteroffensive strategy for interspecies competition.” Molecular microbiology 63.3 (2007): 872-880.…

26. Bao, Xudong, et al. “Streptococcus oligofermentans inhibits Streptococcus mutans in biofilms at both neutral pH and cariogenic conditions.” PloS one 10.6 (2015): e0130962.…

27. Wang, Bing-Yan, and Howard K. Kuramitsu. “Interactions between oral bacteria: inhibition of Streptococcus mutans bacteriocin production by Streptococcus gordonii.” Applied and environmental microbiology 71.1 (2005): 354-362. Inhibition of Streptococcus mutans Bacteriocin Production by Streptococcus gordonii.

28. Van der Hoeven, J. S., and P. J. M. Camp. “Synergistic degradation of mucin by Streptococcus oralis and Streptococcus sanguis in mixed chemostat cultures.” Journal of dental research 70.7 (1991): 1041-1044.…

29. Palmer, Robert J., et al. “Mutualism versus independence: strategies of mixed-species oral biofilms in vitro using saliva as the sole nutrient source.” Infection and immunity 69.9 (2001): 5794-5804. Strategies of Mixed-Species Oral Biofilms In Vitro Using Saliva as the Sole Nutrient Source

30. Grenier, D., and D. Mayrand. “Nutritional relationships between oral bacteria.” Infection and immunity 53.3 (1986): 616-620. Nutritional relationships between oral bacteria.

31. Grenier, D. “Nutritional interactions between two suspected periodontopathogens, Treponema denticola and Porphyromonas gingivalis.” Infection and immunity 60.12 (1992): 5298-5301. Nutritional interactions between two suspected periodontopathogens, Treponema denticola and Porphyromonas gingivalis.

32. Krom, B. P., S. Kidwai, and J. M. Ten Cate. “Candida and Other Fungal Species Forgotten Players of Healthy Oral Microbiota.” Journal of dental research 93.5 (2014): 445-451.…

33. Bang, Corinna, and Ruth A. Schmitz. “Archaea associated with human surfaces: not to be underestimated.” FEMS microbiology reviews (2015): fuv010.…

How far are we from having our bacteria engineered to reduce obesity?


Refers to:

Though we’re still far from knowing enough to precisely engineer human weight loss through gut microbe manipulations, ironically, at least one case study (1) shows it’s possible to influence body weight simply by manipulating gut microbiota. Irony because the outcome in that case turned out to be inadvertent weight gain. Nevertheless this example offers

  • An important proof of principle that microbe transfers could mediate metabolic transformation in humans.
  • Insight into conditions necessary for effecting such transformation.
  • Scientific obstacles that stand in the way of using it to reverse obesity in human subjects.

In this report (1),

  • A 32-year old woman with recurrent Clostridium difficile infection (CDI) underwent Fecal microbiota transplant (FMT), her 16-year old daughter the FMT donor.
  • The FMT proceeded as planned after the daughter was screened and declared clear for HIV1 and 2, syphilis, hepatitis A, B and C, C. difficile, Giardia lamblia and other enteric pathogens,.
  • Post-FMT, the patient’s condition improved with no CDI recurrence . However, 16 months post-FMT, patient reported having gained 34 pounds. She now weighed 170 pounds with a BMI of 33, i.e., obese. This even though she hadn’t initially lost any weight over the months she underwent a series of antibiotic Rx for CDI.
  • Prior to FMT, the patient stably weighed 136 pounds with a BMI of 26 while her daughter weighed ~140 pounds with a BMI of 26.4.

How to explain why this happened?

  • The FMT donor, the patient’s daughter, provides the first clue. 140 pounds at the time of FMT, she too steadily gained weight to subsequently become 170 pounds. Obviously 16 years is the tail end of puberty, a life stage with profound hormonal and other physiological changes, which undoubtedly also impact gut microbiota composition.
  • In fact, the FMT donor’s subsequent weight gain suggests that as she completed puberty, her microbiota were changing in profile from one that supported a lighter weight to one more efficient in harvesting energy and thereby in promoting weight gain.
  • Seen in this light, this FMT recipient’s weight gain merely mirrors that of her donor’s, with the microbiota transfer the most likely change agent. Presumably post-FMT transplant, the recipient’s gut microbiota resembled those of the donor’s, i.e., more efficient in harvesting energy and thereby promoting weight gain.

Colonisation resistance Is An Important Obstacle To Microbe Transplant ‘Take’

While this FMT recipient’s weight gain can be plausibly explained as mimicking that of the donor’s, still the question needs further probing because doing so gets us to the crux of scientific obstacles in using microbe transfers to effect weight changes, be it gain or loss. Crux in this case is gut microbiota status quo in normal individuals and how it contrasts with the situation in this case report.

  • Healthy GI tracts have commensal microbes occupying the various GI tract niches as groups of specialized workers performing essential functions predicated on the needs of the niches they occupy. Thus, regardless a body is lean or obese, healthy guts have or should have niches that are microbially replete and successfully repel not only harmful invaders like C.diff but also other outsiders seeking to occupy the same niche, a process called Colonisation resistance.
  • This FMT recipient’s recurrent CDI suggests her gut microbiota was obviously already in considerable turmoil to start with. Prior to resorting to FMT, her already unstable GI tract microbiota was then subjected to antibiotic Rx consisting of metronidazole, vancomycin, amoxicillin, clarithromycin, rifamixin (1). Thus, by the time she underwent FMT, her GI tract was pretty depleted of stable microbial inhabitants, i.e., of robust colonization resistance capability. The microbes she received from her daughter’s poop would thus have been able to easily colonize the now available niches in her GI tract.
  • Genetics is another important factor, as in relatedness between donor and recipient, which presumably increased the likelihood of ‘take’ of donor’s microbes in the recipient’s GI tract (1). A 2016 study (2) was among the first to a) monitor long-term fate of FMT in human recipients, and b) observe different fates of same-donor poop transplants in different recipients, i.e., some microbial species successfully colonized some recipients but not others. Factors that determine such microbial ‘take” are still not fully clear but likely the two most important are the species and strain fitness within the donor pool on the one hand and colonization resistance in the recipient’s GI tract on the other hand. Clearly, genetic relatedness between donor and recipient is likely to play a role in how similar or different microbiota are to start with between different individuals.

Thus, this case report (1)

  • Provides preliminary proof of principle that microbiota transplant may lead to metabolic transformation in the recipient to mimic those of the donor.
  • Suggests unoccupied or available microbial niches may be a prerequisite for such possibility to convert to actuality.

Some Scientific Obstacles To Successful Weight Loss Through Microbe Transplants In The Obese

  • Unless obese recipients’ GI tracts are prepared prior to microbe transplants to create niches that can accept them, such transplants may not work in reducing obesity. In other words, depleting recipients of their indigenous gut microbes may be a necessary preliminary step. Problem is, antibiotics, easiest tools available to do this, are blunt instruments and each such antibiotic’s effect on indigenous gut microbes will differ from person to person since gut microbial populations vary. Which antibiotics optimally prepare a recipient GI tract for optimal ‘take’ is currently unknown as also whether indigenous microbe-depleting effect of any one antibiotic is even generalizable across different individuals.
    • However, antibiotics may not be the only approach to deplete recipients of indigenous microbes prior to microbe transplant since the 2016 study (2) found bowel lavage alone without prior antibiotic Rx allowed stable donor microbe colonization in FMT recipients. By monitoring recipient gut microbes from 84 up to 400 days, this study showed durable co-existence of some donor and recipient microbial species.
    • Much work still needs to be done to understand how to effect efficient ‘take’ of microbe transplants.
  • Obviously obesity is outcome of diet, microbes and genetics. While interventions such as prior antibiotic Rx and bowel lavage may prepare available GI tract niches for microbial transplants to successfully occupy them initially, can sustainable microbial ‘take’ be assured without more profound, long-term habit changes? Doesn’t the principle of colonization resistance suggest that continuing post-transplant with the same diets that sustained their obesity only increase the likelihood their post-transplant GI niches would continue to preferentially support such obesity-associated microbes? Doesn’t that suggest microbe engineering alone may not suffice unless accompanied by diet change?

Misuse Of A Statistical Tool Is An Obvious Weakness Of Mouse Model Studies Of This Kind

Finally, a couple of points to add to Drew Smith‘s thorough analysis of the study quoted in the question.

One, though mouse is the most prevalent preclinical animal model, the travesty is rarely do its findings translate to humans.

Two, important to note the blatant data manipulation all too common in such mouse studies. While the article in question refers to an as-yet unpublished study presented during an August 2016 conference, this group has published on this NAPE mouse model-associated weight loss in 2014 (reference 6 in Drew Smith’s answer, 3 here). The important bit is in the figure legend. This particular experiment has 4 groups with 10 mice per group. So far so good. Not so good? That the authors chose to show not individual data points for each mouse in each group or mean +/- SD (standard deviation) but rather mean +/- SEM (standard error of the mean).

SEM is derived by dividing the standard deviation (SD) by the square root of the number tested. Let’s say SD in one group was 4.7. By dividing this SD by the square root of 10, the number of mice in the group, one can artificially reduce the variation within this group down to 1.6, i.e., its SEM. Such manipulations make the data look much cleaner and clearly separate the trends between the groups but the fact that these authors had to resort to this gimmick in such a small data set suggests much greater actual variation, i.e., considerable variation within groups and therefore considerable overlap between groups.

Variation is a given in biology especially when experiments involve such complex entities as live animals and human beings. This is compounded by less than optimal precision and accuracy of many biological assays, and widely variable skill and rigor of experimenters. As with any statistical tool, SEM has value when used appropriately as for example when trying to account for inevitable variations between experiments. Repeat the same experiment over time with experimental animals divided into the same 4 groups and there’s likely to be some variation even in the same group across experiments. SEM can help offset such variation and its use in such circumstances is not only appropriate but also tempered by the fact that combining data from different experiments adds more statistical power to the dataset not only by simply increasing the number of subjects per group but also by accounting for inter-experimental variations. That was clearly not the case here. The authors themselves describe this experiment in their paper’s Materials and Methods (3) as one experiment of 40 male mice divided into 4 groups of 10 mice each. In other words, this is misuse of the SEM statistical tool.

If within-group variations are larger than between-group variations, obviously we can’t conclude much especially in small studies where each group has only 10 subjects. Obviously such studies couldn’t get published. With the Publish or perish imperative only strengthening not weakening in recent decades, resorting to Data dredging is also at epidemic proportions. As gatekeepers, scientific peer reviewers and journal editors are responsible for stemming the tide of such abuse of statistics, a factor that also plays an important role in the current biomedicine data irreproducibility crisis. Obviously and dismayingly this example shows that even respectable scientific journals with quite high impact factors like Journal of Clinical Investigation (JCI) still aren’t performing due diligence on the data they choose to publish. No wonder the data irreproducibility crisis shows no sign of abating.


1. Alang, Neha, and Colleen R. Kelly. “Weight gain after fecal microbiota transplantation.” Open forum infectious diseases. Vol. 2. No. 1. Oxford University Press, 2015. Weight Gain After Fecal Microbiota Transplantation

2. Li, Simone S., et al. “Durable coexistence of donor and recipient strains after fecal microbiota transplantation.” Science 352.6285 (2016): 586-589.

3. Chen, Zhongyi, et al. “Incorporation of therapeutically modified bacteria into gut microbiota inhibits obesity.” The Journal of clinical investigation 124.8 (2014): 3391-3406. Incorporation of therapeutically modified bacteria into gut microbiota inhibits obesity