What is the difference between immunotherapy and targeted therapy (oncology)?


Difference between Cancer immunotherapy – Wikipedia and other types of cancer therapies isn’t that the latter alone are targeted since immunotherapy is also targeted therapy, and in fact has the potential to target cancer even more specifically. Rather, the difference is cancer immunotherapy targets cancer indirectly. Other cancer therapies target cancers directly.

Unique features of cancer immunotherapy are

  • Tries to effectively harness the body’s own immune system to eliminate the cancer. Such an effort might be driven by a Monoclonal antibody – Wikipedia (mAb) that targets a checkpoint molecule, i.e., a checkpoint inhibitor, or by re-engineering the patient’s own T cells in vitro to specifically target cancer cells, i.e., CAR-T (Chimeric antigen receptor – Wikipedia) cells, to mention two immunotherapies being studied most intensively.
  • Immunological memory: This hallmark of the Adaptive immune system – Wikipedia could prevent re-appearance of antigenically same/similar cancers and help keep patients relapse-free. Other cancer therapies lack this capacity to prevent future recurrence of antigenically same/similar cancers.
  • Selective cancer targeting: Could better target cancer specifically and spare healthy tissues. It thus has the potential to be a proverbial Magic bullet (medicine) – Wikipedia, i.e., a cancer treatment with minimal collateral damage. Fulfilling this promise requires cancer immunotherapy to target cancer-specific antigens, i.e., antigens expressed only by the cancer and not by other body cells. Though still very much theoretical rather than practical reality, this capability of cancer immunotherapy is not as feasible with other types of cancer therapies.



What is the best way to restore microbiome diversity after antibiotic treatment?

Even as of 2017, one can’t generalize how different types and doses of antibiotics, either individually or in various combinations, affect microbiota diversity in various locations in the human body across variables such as age, ethnicity, gender. This makes figuring out how to restore microbiota diversity after antibiotic exposure even more out of reach.

Only a handful of studies have even examined how specific antibiotics change gut microbiota composition and how long such changes last. Too many variables differ between these studies, all of which examined fecal microbiota composition, which doesn’t directly assess gut microbiota but is a commonly used, tenuous as in mechanistically unclear approximation.

  • Different studies
    • Used different assessment methods.
    • Examined effect of different antibiotics given by different routes, either oral or intravenous.
    • Followed up for different duration.
    • Included small groups (ranging from a low of 6 to a high of 160) of either patients (intravenous) or healthy volunteers (oral) of different ages and ethnicity.
  • Most studies weren’t randomized and didn’t use a placebo control group.
  • Such studies have inherently poor statistical power, making them unsuitable to extrapolate to the general population.

Other variables that prevent generalization include,

  • Antibiotic effects are quite variable as in not only how they affect an individual but also how they affect microbial composition in different niches within the same body (1, 2).
  • Age is also obviously an important factor as consequences of early-life and geriatric antibiotic exposure are quite different from those in other age groups.
  • Effect of different antibiotics vary.
  • Use of a wide variety of antibiotic treatment courses and doses in the general population makes generalizing practically impossible.

So is there any useful takeaway in the existing data? A 2016 review (3) helpfully summarized data from 16 human studies published in various scientific journals between 1984 and 2015. At a minimum, it lists molecular target, class, resistance mechanism and big picture effect on gut (rather fecal) microbiota composition of the most commonly used antibiotics (see below from 3).

  • Note how information for each major antibiotic is based on simply one or two studies, and at most four (for e.g., for ciprofloxacin), i.e., too few studies and number of subjects to be able to draw firm conclusions.
  • Reduced bacterial diversity and reduced abundance of specific organisms are the overarching observations in most of these studies.
  • Studies with healthy volunteers didn’t report symptoms of gastrointestinal distress. This brings up an important parameter to consider, namely, functional redundancy in microbiota (different bacterial groups that perform the same or similar functions) (4). Such lack of symptoms suggests even if an antibiotic knocks some microbiota members for a loop, albeit temporarily, other organisms, unaffected by such Rx and with overlapping functions, can take over so much so GI tract and other functions carry on seemingly as normal.

Implications for otherwise healthy people is that an occasional antibiotic course will likely temporarily reduce bacterial diversity and abundance of specific bacterial groups, that recovery time and scope will vary from person to person, that recovery occurs even in the absence of specific steps in the form of special diets, supplements, pre- and probiotics, etc., and that recovery to pre-Rx baseline becomes less and less likely with greater frequency of antibiotic exposure (5, reference 69 in table above).

Also important to keep in mind as human microbiota studies continue apace that not just antibiotics but many other types of medications also appear to have capacity to reshape human microbiota community composition. Case in point, Metformin – Wikipedia. Of all things, totally unexpectedly, even this old, tried and tested type II diabetes medication was found to modify gut (fecal) microbiota (6), specifically to reverse the depletion of the all-important butyrate-producing classes of bacteria, a depletion that’s now emerging as a hallmark and sign of gut microbiota dysregulation in type II diabetes. Butyrate-producing bacteria influence Short-chain fatty acid – Wikipedia (SCFA) metabolism, begging the question whether metformin alleviates type II diabetes by acting directly on tissues or rather indirectly, by reshaping gut microbiota composition into one more reflective of health than disease.


1. Jakobsson, Hedvig E., et al. “Short-term antibiotic treatment has differing long-term impacts on the human throat and gut microbiome.” PloS one 5.3 (2010): e9836. http://journals.plos.org/plosone…

2. Zaura, Egija, et al. “Same exposure but two radically different responses to antibiotics: resilience of the salivary microbiome versus long-term microbial shifts in feces.” MBio 6.6 (2015): e01693-15. Resilience of the Salivary Microbiome versus Long-Term Microbial Shifts in Feces

3. Langdon, Amy, Nathan Crook, and Gautam Dantas. “The effects of antibiotics on the microbiome throughout development and alternative approaches for therapeutic modulation.” Genome medicine 8.1 (2016): 39. https://genomemedicine.biomedcen…

4. Mikkelsen, Kristian Hallundbaek, Kristine Højgaard Allin, and Filip Krag Knop. “Effect of antibiotics on gut microbiota, glucose metabolism and body weight regulation: a review of the literature.” Diabetes, Obesity and Metabolism 18.5 (2016): 444-453. https://www.researchgate.net/pro…

5. Dethlefsen, Les, and David A. Relman. “Incomplete recovery and individualized responses of the human distal gut microbiota to repeated antibiotic perturbation.” Proceedings of the National Academy of Sciences 108.Supplement 1 (2011): 4554-4561. http://www.pnas.org/content/108/…

6. Forslund, Kristoffer, et al. “Disentangling the effects of type 2 diabetes and metformin on the human gut microbiota.” Nature 528.7581 (2015): 262. https://www.ncbi.nlm.nih.gov/pmc…


Why is a T helper cell called the cluster of differentiation 4 cell?


Cluster of differentiation – Wikipedia or CD is simply a nomenclature convention that started taking shape in the early 1980s to counter an utterly chaotic system of naming cell types by the name of the Monoclonal antibody – Wikipedia (mAb) that bound a cell-surface molecule specifically expressed by them.

The idea of using mAbs tagged with a fluorescent tag to use as probes to specifically identify specific cell subsets away from others was largely the brainchild of the late Alan Frederick Williams – Wikipedia, who pioneered this approach while working with Rodney Robert Porter – Wikipedia. This was shortly after César Milstein – Wikipedia and Georges J. F. Köhler – Wikipedia published their seminal paper on the creation of mAbs in 1975.

Soon labs around the world started developing mAbs using Milstein and Kohler’s method, tagging them with different fluorescent tags and using them to probe the surface of cells to better characterize them. Naturally different groups started naming cell subsets based on the name of the mAb they were using to bind (tag) a specific cell-surface marker expressed on their surface. Chaos soon reigned. Were leu-3 positive cells the same as OKT4 positive ones or not? What about OKT8 positive cells? Were they the same as Ly2+ or not? And so on ad nauseam.

The CD nomenclature system was established to bring order to this unending chaos. As a result, OKT4 positive cells got labeled as CD4 T helper cell – Wikipedia. OKT8 positive cells got labeled as CD8 Cytotoxic T cell – Wikipedia. And so on.

And of course using mAbs tagged with a wide variety of tags as tools in any number of assays is today one of the most standard techniques in all of biology with hardly any practitioner aware of who actually came up with such a simple but unquestionably brilliant idea.


Are mortality rates for pancreatic cancer primarily due to late diagnosis or other factors?


Pancreatic cancer has relatively poor long-term prognosis compared to other cancers (1, 2). Accounting for >90% of all pancreatic cancer, Pancreatic ductal adenocarcinoma (PDAC) is reported to have

  • Only ~18% one year survival rate for all stages of disease, with <4% surviving 5 years (1, 3).
  • Average survival rate of merely 4 to 6 months following diagnosis (4).

Combination of few specific symptoms and rapid tumor progress (1) contribute to late diagnosis of pancreatic cancers, i.e., they’re usually at an advanced stage already at the time of diagnosis. One study (5) on 116 patients with locally advanced pancreatic cancer diagnosed between December 1998 and October 2009 found initial PDAC diagnosis to consist of ~30% with locally advanced and ~50% with metastatic disease.

Apparently only ~10% of these cancers are operable (1), they tend to metastasize early (6) even as they appear to have astounding mutational capacity. For example, one study found each pancreatic tumor cell to carry an average of 63 genetic alterations, mostly point mutations (7), which makes targeting these tumors much more challenging than normal.

Thus, combination of few specific symptoms, rapid tumor progress, low operability, high rates of mutational loads and metastasis circumscribe Rx options and seem to be major contributors to high mortality rates of pancreatic cancer. As a result, while accounting for only ~3% of all new cancer diagnoses, pancreatic cancer was the 4th most common cause of cancer death in the US in 2013 (8).


1. Hidalgo, Manuel, et al. “Addressing the challenges of pancreatic cancer: future directions for improving outcomes.” Pancreatology 15.1 (2015): 8-18. https://www.researchgate.net/pro…

2. Carrato, A., et al. “A systematic review of the burden of pancreatic cancer in Europe: real-world impact on survival, quality of life and costs.” Journal of gastrointestinal cancer 46.3 (2015): 201-211.)https://www.researchgate.net/pro…

3. Löhr, Matthias. “Is it possible to survive pancreatic cancer?.” Nature Reviews. Gastroenterology & Hepatology 3.5 (2006): 236.

4. Siegel, Rebecca, et al. “Cancer statistics, 2014.” CA: a cancer journal for clinicians 64.1 (2014): 9-29. http://onlinelibrary.wiley.com/d…

5. Malik, Nadia K., et al. “Treatment of locally advanced unresectable pancreatic cancer: a 10-year experience.” Journal of gastrointestinal oncology 3.4 (2012): 326. https://www.ncbi.nlm.nih.gov/pmc…

6. Tuveson, David A., and John P. Neoptolemos. “Understanding metastasis in pancreatic cancer: a call for new clinical approaches.” Cell 148.1 (2012): 21-23. http://www.cell.com/cell/pdf/S00…

7. Jones, Siân, et al. “Core signaling pathways in human pancreatic cancers revealed by global genomic analyses.” science 321.5897 (2008): 1801-1806. http://www.gastricbreastcancer.c…

8. Siegel, Rebecca, Deepa Naishadham, and Ahmedin Jemal. “Cancer statistics, 2013.” CA: a cancer journal for clinicians 63.1 (2013): 11-30. http://onlinelibrary.wiley.com/d…


Why can immunotherapy achieve long-term survival? Is long-term survival unique to immunotherapy?

The figure referenced in the question, “why can immunotherapy achieve long-term survival?” is taken from a 2012 review in Clinical Cancer Research (1).

Figures have figure legends for a reason. Without a figure legend it’s hard to decipher how to interpret and what to conclude from a figure. So here’s the same figure with its authors-provided legend (1) providing the explanation and context.

Clearly, the authors are using this figure to argue that when current forms of cancer immunotherapy work, which they do in small subsets of patients, they are able to induce long-term survival and even cure, a scenario far different from therapies that target oncogenes, where the response is initially robust in most patients but long-term survival is still not substantially improved over the comparator (older types of therapies). So essentially the question is why such a difference in long-term outcome between immunotherapy that tries to trigger tumor-specific immune responses versus therapies that directly target the tumor?

  • Checkpoint inhibitors are among the most widely tested Cancer immunotherapy – Wikipedia Rx thus far. While not every patient on checkpoint inhibitors is seen to benefit from them in clinical trials, a small, variable fraction (~10%) appears to gain long-term benefit in the form of a consistent, absolute increase in survival at the end of years-long follow-up. Much of this survival is seen to be both progression- and relapse-free.
    • While checkpoint blockade doesn’t directly target tumors (2, 3, 4), its goal is to neutralize a commonly used approach of many tumors to evade tumor-specific immune responses, namely, by expressing cell-surface molecules which can bind their ligands on tumor-specific T cells and inhibit them. Such neutralization of tumor evasion strategies would then make already existing tumor-specific T cells effective in getting rid of the tumor.
  • Therapies such as those that target oncogenes directly target the tumor and attack it to rid of it, with some degree of collateral damage of course. But that’s it. Such tumor-targeting therapies lack capacity to prevent new tumors from arising, nor can they re-direct their attack on resistant variants that arise from the tumor mass in response to their attack. After all tumors are a microcosm of real-time evolution. The right hand graph in the figure above thus represents rapid initial success in controlling the tumor but no major benefit in survival improvement over the long-term.

So what could account for this difference in long-term survival, albeit thus far in only a handful of patients treated with checkpoint inhibitors?

Briefly, immunological memory. In those patients where checkpoint blockade successfully frees up tumor-specific T cells to get fully activated by tumor antigen(s), such cells then attack tumor cells and kill them. Ensuing reduction of tumor mass and corresponding reduction of tumor antigen(s) would then drive such cells to differentiate into memory cells, a hallmark of the Adaptive immune system – Wikipedia. Such memory cells would precisely be poised to get re-activated and kill any newly arising tumor cells in future, similar to what happens post-infection for example. This could explain cancer immunotherapy’s capability to induce long-term progression- and relapse-free survival, albeit currently in small sets of clinical trial patients.

As for tumor escape variants, those should still remain a possibility even with checkpoint blockade, unless the way in which tumor-specific immune cells engage tumors somehow intrinsically reduces scope and chance for tumor resistance (5). That is very much speculative at the moment since mechanism of tumor control following checkpoint blockade is not all that clear at the moment. For example, while Programmed cell death protein 1 – Wikipedia (PD-1) blockade appears capable of expanding memory T cells within tumors (6), clinical response by blockade using anti- CTLA-4 – Wikipedia (Ipilimumab – Wikipedia) doesn’t necessarily correlate with memory T cell changes (7).


1. Ribas, Antoni, et al. “New challenges in endpoints for drug development in advanced melanoma.” Clinical Cancer Research 18.2 (2012): 336-341. http://clincancerres.aacrjournal…

2. Tirumalai Kamala’s answer to What experiments can we do with TILs to work out the mechanism of action of checkpoint inhibitors?

3. Tirumalai Kamala’s answer to Why do some studies use CMV recall assays to check in vitro functional activity of checkpoint inhibitors?

4. Tirumalai Kamala’s answer to How does PD-L1 checkpoint inhibition selectively target cancer cells but not healthy cells?

5. Tumeh, Paul C., et al. “PD-1 blockade induces responses by inhibiting adaptive immune resistance.” Nature 515.7528 (2014): 568. https://www.ncbi.nlm.nih.gov/pmc…

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

7. Yuan, Jianda, et al. “Novel technologies and emerging biomarkers for personalized cancer immunotherapy.” Journal for immunotherapy of cancer 4.1 (2016): 3. https://jitc.biomedcentral.com/t…


Why do some IgG antibodies give protection and others don’t?


The immune system is capable of making many different types of immune responses to the same antigen, some of which will be effective, others less so and still others ineffective. Outcome of a given immune response usually reflects the balance of all these various types of immune responses that constitute it.

IgG antibodies are usually very effective against some types of bacteria such as Polysaccharide encapsulated bacteria – Wikipedia whose examples include pathogens such as Haemophilus influenzae – Wikipedia, Streptococcus pneumoniae – Wikipedia, Neisseria meningitidis – Wikipedia.

Usually extracellular, these bacteria multiply outside of body cells. Such bacteria can be dangerous, even life-threatening if they gain access to internal organs and tissues such as the brain, from their initial portals of entry such as nasal/oral passages. In such situations, presence of sufficient titers of bacterial antigen(s)-specific IgG can stop the spread of these bacteria dead in their tracks by neutralizing not just such antigens, a very important disease-alleviating function when such antigens are toxins, but also the whole bacteria themselves. This is also why transfer of such IgG antibodies can be protective, for example in the form of maternal IgG in the case of Passive immunity – Wikipedia.

OTOH, IgG antibodies can be similarly antigen-specific but just not effective when the source of their antigen is an intracellular organism that might be spreading stealthily from cell to cell through membrane-coated vesicles, for example.

Thus, like any other type of adaptive immune response, IgG antibodies are directly effective if ultimately their effector function can directly reduce the source of their antigen. Since, however, no single entity within the immune system can ‘see’ the ‘whole’ target, be it virus, bacteria, allergen, tumor cell or even transplant tissue, over the course of its evolution, the system appears to have chosen to hedge its bets and allow a broad range of different types of immune responses to develop over the course of a process, with the intent that at least one of them would prove effective, an approach that, given how our species is flourishing in terms of sheer numbers, suggests seems to have worked spectacularly well.


Is the healthy human brain home of a microbiome?

Does the healthy human brain have its own stable microbiota? Thus far, at least one peer-reviewed study (1) appears to suggest ‘healthy’ human brains could have their own share of microbiota.

A simple enough study (1), the authors performed deep sequencing of white matter-derived RNA from 4 HIV patients, 4 other disease controls and 2 cerebral surgical resections from epilepsy patients, and found alpha-proteobacteria in all of them as well as some herpes viruses and bacteriophages. Given none of the brains were technically from ‘healthy’ individuals, results from such a small study are inconclusive and will remain so until replicated or until other studies report finding other micro-organisms in healthy human brains. However, such a task is very much uphill and definitely not for the faint-hearted.

Human gut microbiota is unremarkable because it’s entirely expected. OTOH, brain microbiota is far more controversial simply because microbes are unexpected in such supposedly sterile organs as the healthy brain. Any and every critique applied to human brain microbiota is just as applicable to human gut microbiota since similar methods are used to analyze microbiota anywhere.

After all, microbiota analysis methods have considerable problems in the form of study design flaws, poor data quality and reproducibility, and ambiguous and questionable statistical approaches used for data analysis (2, 3, 4, 5, 6, 7, 8), except such critiques are usually glossed over in human gut microbiota studies even as they would likely be the centerpiece of focus about microbiota found in unexpected places such as the human brain.

Finding microbes in the healthy human brain would thus get parsed using much more stringent critiques that just don’t get applied to human gut microbiota analysis. Unsurprising then that the result of this sole study was easily considered suspect and discredited with the argument that microbiota such as alpha-proteobacteria are often found as contaminants (9). Casting doubt on the method casts doubt on the results.

Rightly or wrongly, the brain is typically privileged as the seat of the super-self. Microbes in the healthy brain necessitates some means of surveillance by the immune system but for a mindset that privileges the brain, both microbes and the immune system are potentially threatening invaders, given their unpredictable capacity for invasion, damage and just plain downright mischief. No surprise then that for long the consensus has been that the brain has systems in place to keep both the immune system and microbes at bay.

One of the clearest examples of such a system is the notion of Immune privilege – Wikipedia, an idea initially articulated by Peter Medawar – Wikipedia, which holds that being exquisitely vulnerable to irreversible inflammatory damage, certain parts of the body such as the brain limit their interaction with the environment beyond, and with the immune system in particular. The Blood–brain barrier – Wikipedia (BBB) embodies immune privilege in the form of a physical barrier preventing the immune system from fully accessing the brain. If the circulating immune system itself is supposed to have limited access to the brain, it’s not surprising that it then follows that microbes in the brain would be considered not a feature of health but of disease.

However, the notion of immune privilege lacks evolutionary coherence. Any part of the body inaccessible or poorly or less accessible to the immune system renders it unprotected. Akin to open sesame to pathogens, such an idea makes no evolutionary sense whatsoever. All the more flabbergasting then that this idea remained an entirely acceptable construct for decades and lingers on even in current thinking, as evidenced by an entire Wikipedia article devoted to it that discusses it wholly at face value, without ever referring to the hugely problematic implication of its evolutionary unworkability.

Steady drumbeat of data in recent years has however provided a couple of countervailing pieces of evidence that suggest the ground may be fast getting cut out from under the bastions holding the old ideas of immune privilege and BBB in place,

  • Numerous studies (see reviews 10, 11, 12, 13, 14), especially in preclinical animal models, have now shown gut microbiota influence many if not most aspects of brain development and function.
  • The 2015 report that contrary to long-held dogma, the (rodent) brain isn’t devoid of lymphatics (15, 16, 17). The so-called BBB may not be an unimpregnable barricade after all.

There remains then only the small (ahem!) matter of reconciling current immune system theory to healthy human brain microbiota as in how could the immune system tolerate them without constant, protracted inflammation? It’s doable if one accepts that any commensal microbiota that applied evolutionary selection pressure would be tolerated through the process of thymic Central tolerance – Wikipedia wedded to antigen-specific Regulatory T cell – Wikipedia development and function.

However, current immunological dogma would consider such an argument heretical since it assumes the immune system ontogenetically learns within each individual’s lifetime to distinguish what is self from what is not. OTOH, microbiota in the healthy human brain is not a problem for those like me who believe immune, especially T cell, development to be a phylogenetically powered process that is being perfected over the evolutionary history of a species. Only time and weight of scientific evidence favoring one or the other view will settle that debate.


1. Branton, William G., et al. “Brain microbial populations in HIV/AIDS: α-proteobacteria predominate independent of host immune status.” PloS one 8.1 (2013): e54673. http://journals.plos.org/plosone…

2. Lozupone, Catherine A., et al. “Meta-analyses of studies of the human microbiota.” Genome research 23.10 (2013): 1704-1714. Meta-analyses of studies of the human microbiota

3. Goodrich, Julia K., et al. “Conducting a microbiome study.” Cell 158.2 (2014): 250-262. http://ac.els-cdn.com/S009286741…

4. McMurdie, Paul J., and Susan Holmes. “Waste not, want not: why rarefying microbiome data is inadmissible.” PLoS computational biology 10.4 (2014): e1003531. http://journals.plos.org/ploscom…

5. Sinha, Rashmi, et al. “The microbiome quality control project: baseline study design and future directions.” Genome biology 16.1 (2015): 276. https://genomebiology.biomedcent…

6. Weiss, Sophie, et al. “Correlation detection strategies in microbial data sets vary widely in sensitivity and precision.” ISME J 10.7 (2016): 1669-1691. https://www.researchgate.net/pro…

7. Boers, Stefan A., Ruud Jansen, and John P. Hays. “Suddenly everyone is a microbiota specialist.” Clinical Microbiology and Infection 22.7 (2016): 581-582. http://www.clinicalmicrobiologya…

8. Bik, Elisabeth M. “Focus: microbiome: the hoops, hopes, and hypes of human microbiome research.” The Yale journal of biology and medicine 89.3 (2016): 363. https://www.ncbi.nlm.nih.gov/pmc…

9. Salter, Susannah J., et al. “Reagent and laboratory contamination can critically impact sequence-based microbiome analyses.” BMC biology 12.1 (2014): 87. https://bmcbiol.biomedcentral.co…

10. Diamond, Betty, et al. “It takes guts to grow a brain.” Bioessays 33.8 (2011): 588-591. https://www.researchgate.net/pro…

11. Al-Asmakh, Maha, et al. “Gut microbial communities modulating brain development and function.” Gut microbes 3.4 (2012): 366-373. http://www.tandfonline.com/doi/p…

12. Collins, Stephen M., Michael Surette, and Premysl Bercik. “The interplay between the intestinal microbiota and the brain.” Nature reviews. Microbiology 10.11 (2012): 735.

13. Tillisch, Kirsten. “The effects of gut microbiota on CNS function in humans.” Gut microbes 5.3 (2014): 404-410. https://www.ncbi.nlm.nih.gov/pmc…

14. Sampson, Timothy R., and Sarkis K. Mazmanian. “Control of brain development, function, and behavior by the microbiome.” Cell host & microbe 17.5 (2015): 565-576. http://ac.els-cdn.com/S193131281…

15. Louveau, Antoine, et al. “Structural and functional features of central nervous system lymphatics.” Nature 523.7560 (2015): 337. https://www.researchgate.net/pro…

16. Tirumalai Kamala’s answer to Why aren’t there lymph nodes in the brain?

17. Tirumalai Kamala’s answer to Why did it take so long to discover that the brain is connected to the immune system?


Each B cell is antigen specific. How many such B cells would be present for a particular antigen in the whole body?

It’s difficult enough to estimate the total number of B cells in the body, let alone the number of B cells specific for any given antigen, i.e., antigen-specific B cell frequency, aka precursor frequency. Additional obstacles include the fact that the pool of cells being analyzed include

  • B cells at various stages of development, especially in the bone marrow.
  • Not just conventional adaptive but also innate B cell subsets.
  • Not just naive (antigen-inexperienced) but also memory B cells.

Though they all express antigen-specific receptors, the B-cell receptor – Wikipedia (BCR), which when secreted is called the antibody, B cells aren’t a monolithic entity. Rather, the classical B (as also T) cell subsets with somatically recombined antigen receptors (V(D)J recombination – Wikipedia) belong to the adaptive immune system, which is characterized by remarkable diversity. Such classical or conventional B cells are B-2 or Follicular B cells. They constitute the bulk of B cells in the lymph nodes, spleen, bone marrow and in circulation.

However, other B cell subsets such as B-1 cell – Wikipedia and Marginal zone B-cell – Wikipedia (MZ B) also secrete antibodies, mostly IgM, some IgG3, usually also termed Natural antibodies – Wikipedia, but they tend to not somatically recombine their antigen receptors, i.e., they retain germline receptors, to not circulate, to not convert into memory cells, and to perform their effector functions without the help of T cells.

Thus, frequency of a given antigen-specific B cell is obviously very different between the ‘innate’ and ‘adaptive’ subsets of B cells. Greater the receptor diversity, lower the frequency of any specificity, simply for a practical reason, there just isn’t enough space in the body to house expanded numbers of each and every antigenic specificity, given that the total B cell pool in the body is estimated to be 1 to 2 X 10^11 (1, 2; also see below from 3).

Frequency of individual B cell specificities in humans are also exceedingly difficult to estimate simply because of difficulty of access to source material. Blood, an obvious choice to sample, is estimated to harbor only ~2% of the total B cell pool (4, also see above from 3), with much higher numbers present in lymph nodes (~28%), spleen (~23%) and red bone marrow (~17%) (3, 4).

Add the additional complexity that humans are estimated to have ~600 to 750 lymph nodes (3, 5, 6, 7) and difficulty of enumerating antigen-specific B cell frequency becomes more than obvious. Further, even a mere ~2% of total body B cell numbers still amounts to ~ 2 to 4 X 10^9 B cells in blood. Reasonably accurate estimates crucially turn on how much needs to be minimally sampled and method(s) used to sequence and quantitate BCRs (see below from 8, emphasis mine).

The consequences of insufficient biological sampling have been investigated previously by Warren and colleagues [26]: they showed that distinct 20 ml blood samples from the same individual captured only a portion of the TCR peripheral blood repertoire (biological undersampling). Furthermore, technological undersampling has been shown to compromise the detection of ‘public’ clones (clones shared across individuals), which are a common target in immune repertoire studies [27,28]. In fact, several studies indicated that there was a positive correlation between sequencing depth and the number of public clones detected [13,29,30]. Thus, the biological conclusiveness of a study benefits from the implementation of biological replicates (test for biological undersampling [26,31,32]) (Figure 1A) and technical replicates (test for technological undersampling [33–36]) (Figure 1A), which may be performed once for each cell population analyzed. It is important to note that biological undersampling can only be meaningfully addressed if sufficient technological sampling has been established [33]. Furthermore, species accumulation and rarefaction analyses may be performed to quantify the extent of (under)sampling [29,33,35,37]

Thus, there is likely to be substantial margin of error in estimates based on blood B cells. As well, antigen-specific B cell frequency is unsurprisingly extremely dynamic, changing with age and unpredictably varying exposure to antigens over time.

For what it’s worth, a commonly bandied about estimate of antigen-specific B cell frequency in the circulating, naive repertoire is one in 10^5 to 10^6. Extrapolating from blood and totally disregarding the contribution of memory and innate B cells to the estimated total B cell number of 1 to 2 X 10^11, that means each B cell specificity could range from 1 to 2 X 10^5 to 10^6 as also a total of 1 to 2 X 10^5 to 10^6 different individual B cell specificities or capacity to bind that many different antigens. Obviously, since memory and innate B cells are indeed part of the total B cell number, actual numbers of naive B cells specific for a particular antigen are likely markedly lower.

Taking such estimates at face value, is that sufficient frequency and diversity, given that over the course of a lifetime an anticipatory defense system such as the B cell has to contend with a potential universe of antigens that is likely orders of magnitude higher? Important at this point to recall that in B cells, the naive or antigen-inexperienced repertoire diversity is bolstered, maybe even more than amply so, by three other cardinal features, namely, clonal proliferation, Cross-reactivity – Wikipedia (which some refer to as polyreactivity) and Somatic hypermutation – Wikipedia (SHM), with that last, SHM, being a unique property of conventional B cells.

  • Clonal proliferation is the capacity of an activated B or T cell to quickly proliferate (some estimates even suggest dividing every 16 to 20 hours). Progeny of each such single B or T cell are their clones having the exact same antigen receptor (BCR or TCR). Thus, at the height of an immune response, say during the acute phase of an infection, frequencies of individual B or T cells could increase to as many as 1 in 10^3 to 1 in 10^4, i.e., a 100 to 1000-fold increase, at least locally. Clonal proliferation helps bolster the sufficiency of antigen-specific B cell frequency.
  • Cross-reactivity (aka polyreactivity) is the capacity for a given BCR (and antibody) to bind more than one antigen. Often but not always, this relates to structural similarity between different antigens. After all, though the antigenic universe is vast, biology still dictates its sequence and structural constraints. Cross-reactivity helps mitigate the potential insufficiency of antigen-specific B cell diversity.
  • SHM is the process by which conventional B cells that bound their specific antigen and presented pieces of it in the MHC (Major histocompatibility complex – Wikipedia) to cognate T cells receive T cell help that drives mutations within the V gene segment of the BCR. This creates BCR (and antibody) variants additional to those generated during primary B cell development by somatic recombination. Thus, though monozygotic (identical) twins have nearly identical primary antibody repertoire (9), meaning it is largely the product of genetic factors, antigenic experience over time, which can be highly individual and variable, leads to a secondary repertoire that can vary substantially between individuals, even identical twins. SHM helps enhance antigen-specific B cell diversity.


1. Morbach, H., et al. “Reference values for B cell subpopulations from infancy to adulthood.” Clinical & Experimental Immunology 162.2 (2010): 271-279. Reference values for B cell subpopulations from infancy to adulthood

2. Greiff, Victor, et al. “Bioinformatic and statistical analysis of adaptive immune repertoires.” Trends in immunology 36.11 (2015): 738-749. https://www.researchgate.net/pro…

3. Apostoaei, A. Iulian, and John R. Trabalka. “Review, Synthesis, and Application of Information on the Human Lymphatic System to Radiation Dosimetry for Chronic Lymphocytic Leukemia.” Inc., Tennessee (2012). https://www.cdc.gov/NIOSH/ocas/p…

4. Georgiou, George, et al. “The promise and challenge of high-throughput sequencing of the antibody repertoire.” Nature biotechnology 32.2 (2014): 158-168. https://www.researchgate.net/pro…

5. Trepel, F. “Number and distribution of lymphocytes in man. A critical analysis.” Klinische Wochenschrift 52.11 (1974): 511-515.

6. Valentin, Jack. “Basic anatomical and physiological data for use in radiological protection: reference values: ICRP Publication 89.” Annals of the ICRP 32.3 (2002): 1-277. http://dspace.elib.ntt.edu.vn/ds…

7. Agur, Anne MR, and Arthur F. Dalley. Grant’s atlas of anatomy. Lippincott Williams & Wilkins, 2009.

8. Greiff, Victor, et al. “Bioinformatic and statistical analysis of adaptive immune repertoires.” Trends in immunology 36.11 (2015): 738-749. https://www.researchgate.net/pro…

9. Glanville, Jacob, et al. “Naive antibody gene-segment frequencies are heritable and unaltered by chronic lymphocyte ablation.” Proceedings of the National Academy of Sciences 108.50 (2011): 20066-20071. http://www.pnas.org/content/108/…


How does PD-L1 checkpoint inhibition selectively target cancer cells but not healthy cells?


, ,

PD-L1 – Wikipedia checkpoint inhibition doesn’t selectively target cancer cells. Rather, targeting PD-L1 with a PD-L1-specific monoclonal antibody (mAb) prevents it from engaging with its ligand, PD-1 (Programmed cell death protein 1 – Wikipedia), a cell-surface molecule largely, though not exclusively, expressed on the surface of immune cells such as T (both CD4 and CD8), B, NK (Natural Killer) T cells, monocytes and some DCs (dendritic cells) (1, 2).

PD-L1 and PD-L2 are expressed on the cell-surface of a much wider variety of cells, being reported not just on T and B cells but also endothelial and epithelial cells, heart, lung, skeletal muscle, placenta, among others (2, 3). PD-L1 became relevant for cancers when multiple studies reported (3)

  • Its high level expression by many types of cancers such as Breast, Cervix, Colon, Esophagus, Liver, Lung, Kidney, Ovary, Pancreas, Skin.
  • Its expression by tumors correlated with poorer patient prognosis.

At the same time, multiple studies also correlated high PD-1 expression levels on Tumor-infiltrating lymphocytes – Wikipedia (TILs) with poor prognosis of cancer patients as well as poor effector function (anti-tumor activity) of such TILs in in vitro studies (3).

High PD-L1 expression on tumor cells is considered a tumor adaptation attempting to thwart effective anti-tumor immune responses by inhibiting PD-1-expressing TILs. Persistent T cell expression of PD-1 is interpreted as a sign of T cell exhaustion, a colorful description signifying the cell is or has become poorly capable of performing its antigen-specific effector functions.

  • In the case of helper CD4 T cells, PD-1 expression implies poor capacity to help B and cytotoxic CD8 T cells perform their effector functions.
  • In the case of cytotoxic CD8 T cells, PD-1 expression implies poor capacity to kill their target cells.

The hope behind PD-L1 or PD-1 blockade is doing so would release from inhibition PD-1-expressing cancer-specific T cells present in the tumor (and maybe even anywhere else in the body), and thus render them capable of attacking and ridding of the tumor since blocking PD-1PD-L1 engagement was found to reverse lymphocyte effector function inhibition, at least in preclinical (mouse model) studies.

Ideally, the most optimal cancer immunotherapy approach would be cancer antigen-specific since they would likely be those with minimal collateral cost. For example, where an immune cell, say a cytotoxic CD8 T cell specific for a cancer cell antigen, bound its target antigen on the surface of a cancer cell and killed it.

Obviously, PD-L1 or PD-1 blockade is a very different process, affecting not just tumor antigen-specific lymphocytes but others as well so it’s not surprising to note then that it specifically and checkpoint inhibitors in general have at least two major drawbacks.

  • They are not antigen-specific in the strict immunological sense, i.e., they do not target an antigen expressed only by the tumor but not by healthy cells. Thus there is scope for off-target effects (4), meaning attack on non-tumor tissue(s) as well. The hope there is that careful application of blockade dose and frequency would help focus the Rx more to cancer cells and help mitigate targeting of healthy tissue cells.
  • Tumor-infiltrating and therefore presumably tumor-specific T cells could express not just PD-1 but multiple cell-surface inhibitory receptors such as LAG3 – Wikipedia (5) and TIM-3 (HAVCR2 – Wikipedia) (6). Blocking PD-1 alone on such T cells might not suffice to reverse their inhibition. May need to block these other inhibitory molecules as well.

PD-L1 blockade also suffers from an additional drawback, namely, the lack of reliable identification, which means lack of reliable targeting. Identification of PD-L1 expressing cells was mired in technical difficulties since antibodies specific for human PD-L1 had a track record of poor validation. This made it hard to accurately and reliably ascertain whether a particular tumor sample expresses PD-L1 or not. This improved only in recent years after technically validated Immunohistochemistry – Wikipedia (IHC) assays using specific anti-human PD-L1 antibody clones such as Dako/BMS clone 28-8, Merck’s mAb clone 22C3, and Ventana (Genentech/Roche) mAb clone SP142 appeared on the scene (7, 8).


1. Ishida, Yasumasa, et al. “Induced expression of PD-1, a novel member of the immunoglobulin gene superfamily, upon programmed cell death.” The EMBO journal 11.11 (1992): 3887. https://www.ncbi.nlm.nih.gov/pmc…

2. Keir, Mary E., et al. “PD-1 and its ligands in tolerance and immunity.” Annu. Rev. Immunol. 26 (2008): 677-704.

3. Ohaegbulam, Kim C., et al. “Human cancer immunotherapy with antibodies to the PD-1 and PD-L1 pathway.” Trends in molecular medicine 21.1 (2015): 24-33. https://pdfs.semanticscholar.org…

4. Fay, André P., et al. “The management of immune-related adverse events associated with immune checkpoint blockade.” Expert Review of Quality of Life in Cancer Care 1.1 (2016): 89-97. http://www.tandfonline.com/doi/p…

5. Matsuzaki, Junko, et al. “Tumor-infiltrating NY-ESO-1–specific CD8+ T cells are negatively regulated by LAG-3 and PD-1 in human ovarian cancer.” Proceedings of the National Academy of Sciences 107.17 (2010): 7875-7880. http://www.pnas.org/content/107/…

6. Du, Wenwen, et al. “TIM-3 as a Target for Cancer Immunotherapy and Mechanisms of Action.” International journal of molecular sciences 18.3 (2017): 645. TIM-3 as a Target for Cancer Immunotherapy and Mechanisms of Action

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

8. Gandini, Sara, Daniela Massi, and Mario Mandalà. “PD-L1 expression in cancer patients receiving anti PD-1/PD-L1 antibodies: A systematic review and meta-analysis.” Critical reviews in oncology/hematology 100 (2016): 88-98. https://www.researchgate.net/pro…


Is there any scientific proof that vaccines cause autism?

This answer briefly summarizes some overarching inferences,

  • Autism – Wikipedia / Autism spectrum – Wikipedia (Autism Spectrum Disorders, ASD) rates started greatly increasing in some countries such as the US and UK since the 1980s even as doctors little understood these conditions and offered little of value to increasingly anxious parents desperately seeking definitive answers. Thus, in such an Autism causation vacuum, Andrew Wakefield – Wikipedia et al’s 1998 Lancet report (1), the first to offer an explanation for the ‘autism epidemic’, became a convenient crutch for many frustrated parents who felt either ignored or condescended to by the medical establishment.
  • However, in ~20 years, there’s surprisingly scant scientific evidence to support the contention that ‘vaccines cause autism’. Surprising because 20 years is a long enough period to be able to bolster the argument with solid data sets.
  • Even taken at face value, many risk factors about Autism/ASD simply cannot be explained by a ‘vaccines cause autism’ notion. The more facts it can explain about a given phenomenon, the stronger a given hypothesis. That is just not the case with the ‘vaccines cause autism’ notion, which is simply inherently scientifically weak.

On a topic so controversial as a potential vaccine(s)-Autism link, it may be best to start by scrutinizing the original data that got this particular idea started. In 1998, Andrew Wakefield and 12 co-authors published a Lancet article on 12 children, claiming they had identified in them evidence of a novel syndrome they called Autistic enterocolitis – Wikipedia (1).

To digress just a bit at first, it’s somewhat surprising that there isn’t yet an agreed-upon consensus on the etiquette regarding scientific papers that have been retracted (2, 3). Specifically, should they continue to be cited in the literature or not? For example, this Wakefield et al paper continues being cited, 85 times already over six months in 2017 according to Google Scholar.

This answer however requires not just citing this paper but also looking at what it actually says since it subsequently served as the launchpad for a purported vaccine(s)-Autism link. While 8 of these 12 children (67%) had received the MMR vaccine by the time of their symptom onset, the authors concluded (see below from 1, emphasis mine),

‘We identify associated gastrointestinal disease and developmental regression in a group of previously normal children, which was generally associated in time with possible environmental triggersWe did not prove an association between measles, mumps, and rubella vaccine and the syndrome described…If there is a causal link between measles, mumps, and rubella vaccine and this syndrome, a rising incidence might be anticipated after the introduction of this vaccine in the UK in 1988. Published evidence is inadequate to show whether there is a change in incidence22 or a link with measles, mumps, and rubella vaccine.23′

Since these authors did suggest ‘a rising incidence [of their newly coined syndrome] might be anticipated after the introduction of this [MMR] vaccine in the UK in 1988‘, if we give them the benefit of the doubt and assume their autistic enterocolitis concords to some extent with Autism, what do epidemiological data show so far? In a nutshell, nothing that supports their supposition. On the contrary, such studies haven’t found a link between vaccines and Autism.

  • A 1999 study of children in North Thames, London, found rising cases of ASD since 1979 without a sharp increase after MMR was introduced in 1988 (4).
  • A 2001 British study found that while Autism rates in 2 to 5 year olds had increased from 8 boys per 10000 to 29, a 3.6-fold increase, from 1988 to 1993, rates of MMR vaccination had remained stable across these birth cohorts, meaning it wasn’t possible to attribute the Autism rate increase to the MMR vaccine (5).

Thus an examination of the original paper that jump-started the vaccines-Autism controversy finds it did not even make such an assertion and that subsequent studies found no evidence of such a link either. OTOH, one detailed review after another has since found that the MMR vaccine

  • Is safe (6, 7).
  • Is unlinked to Autism (8).

The furore, notoriety and controversy about a link between vaccines and Autism begins with this one study of a mere 12 children, only 8 of whom had received the MMR vaccine by the time of their symptom onset, and it turns out the study didn’t even make that claim. Also more accurately, the paper explores not a link between vaccines and Autism in general but rather one specifically between the MMR vaccine and autistic enterocolitis, a syndrome that isn’t listed in medical textbooks.

So, how did a link between vaccines and Autism even get made? Turns out to have been a subsequent interpretation (3), perhaps helped along by an immediate press conference when this paper was published followed by copious contemporaneous sensationalist front-page coverage by several British newspapers (9) of a kind that suggests (3) many couldn’t even be bothered to read what was actually in the paper.

Subsequent uncovering of undisclosed conflicts of interest behind Wakefield’s study followed by predictable establishment backlash against him then cast him in the potent ‘martyr’ mode, which further solidified and enhanced his reputation among parents desperately seeking definitive answers to their children’s Autism/ASD diagnosis, and who also felt Wakefield took them seriously while feeling the medical establishment didn’t (9).

How Autism’s Causation Vacuum was Fertile Soil for Wakefield’s Vaccine-Autism Supposition to take Root

On the face of it, it seems astounding that one small study on 12 patients should have had such an outsize impact. And yet, maybe not so surprising from a sociological perspective. At the time the Wakefield et al paper came out, Autism/ASD rates had already been spiking for several years with no satisfactory explanation from the medical establishment. Perhaps unwittingly, this state of affairs helped stoke and sustain this particular controversy.

  • Autism diagnosis remains the purview of behavioral scientists who base the diagnosis on a highly subjective checklist, not an impartial, objective, quantitative diagnostic test.
  • Even as they tweaked and improved their diagnostic toolkit, which in turn led to increasing rates of diagnosis, doctors had no clear answer for why steadily increasing numbers of children were being diagnosed with Autism from the 1980s, especially in the US and UK.
  • Still little understood, neither reliable objective diagnosis nor specific treatment, let alone cure, yet existed for Autism/ASD, a situation little changed in the years since.
  • With increasing numbers of parents desperately seeking answers to their children’s predicament, a causation vacuum concerning Autism was precisely calamitous and in hindsight, the Wakefield paper appears to have arrived at just the right moment to fill it with something that no one had proposed thus far, a ‘medical explanation for the autism epidemic‘ (see below from 9, emphasis mine).

‘However, the fact that there was no other reported or known reason for the ‘epidemic’ did not exactly help matters. Whatever their overall validity, vaccine hypotheses did plug a gaping hole in scientific knowledge about this condition that everyone thought had been measured so precisely and accurately with a wealth of new measurement tools and scales. How could it be that no one actually knew why autism was increasing?…Wakefield’s work was so popular because it promised so much. It promised to fully explain the autism epidemic, thus it was particularly ironic that epidemiological sciences never supported his claims.’

  • Autism/ASD having historically been and tending to remain the purview of behavioral scientists may, in the grand scheme of things, turn out to have been a major stumbling block that stymied accelerated understanding of these conditions.
  • Ironically, by highlighting gastrointestinal issues in autistic children, Wakefield may have done Autism/ASD research a huge service. After all, ~20 years on, the gut microbiota-brain link is so much better appreciated now and indeed gut Dysbiosis – Wikipedia is today well-recognized as a cardinal feature in substantial numbers of Autism/ASD patients (10, 11).
  • There was and is an urgent need for a more multi-disciplinary approach for both research and diagnosis in the Autism/ASD field. Gastroenterologists, immunologists, microbiologists, geneticists and other specialists would only help not impede better understanding of these conditions by helping develop more scientifically robust diagnostic approaches and helping tailor more targeted therapies.
  • Even in 2017, such cross-disciplinary research on Autism/ASD is sorely lacking. A simple literature search is a clear indication of this. My search for ‘Autism’ in both Nature Reviews Immunology and Nature Reviews Microbiology together turned up a total of only 24 articles, 2001-2017 (12), only 19 in Nature Reviews Gastroenterology and Hepatology, though through 2006-2017, which suggests the gut-microbiota-brain axis is becoming a bigger focus of research (13), while the same search in Nature Reviews Neuroscience turned up almost 10X higher articles (219), 2001-2017 (14). For context, the Nature Reviews series are typically considered among the most influential science review journals for various subjects.
  • History also suggests the Wakefield idea fills the Autism/ASD causation vacuum rather like a square peg in a round hole. After all, it is inherently scientifically weak since there are so many Autism/ASD risk factors that effects of vaccines, adverse or otherwise, simply cannot explain.

So many Autism/ASD Risk Factors that Vaccines can’t explain

How could vaccines possibly explain

  • Why Autism/ASD is more common in boys than girls, ranging from ~4:1 in the 1990s (15) to ~9:1 by the 2010s (16, 17, 18)? If vaccines ’cause’ autism, a person’s gender shouldn’t matter.
  • Why Autism/ASD rates are so much higher in monozygotic (identical) (70-90% concordance) compared to dizygotic (fraternal) (0-30% concordance) twins (19, 20, 21)?
    • Found in disparate populations such as in the UK (22) as well as in Scandinavia (23).
    • Monozygotic twin concordance for autism is a long-standing feature, being observed right from the 1970s in pioneering studies by Michael Rutter – Wikipedia (24).
    • Autism thus has an unmistakably strong genetic component (22), something that could not be explained by environmental factors alone such as effects of vaccine(s), adverse or otherwise.
    • If vaccines ’cause’ autism, a person’s genetic background shouldn’t matter.
  • Autism/ASD connection with maternal and child antibiotic use reported in several studies (25, 26, 27)? This alludes to a different environmental trigger, namely, changes in gut microbiota composition.
  • Consistently identified Autism risk factors such as exposure to traffic-related air pollutants, increased parental age, maternal obesity, diabetes and folic acid deficiency, prenatal viral infection, C-section, preterm birth, low birth weight, limited or absent breastfeeding, abnormal melatonin synthesis, hyperbilirubinemia, zinc deficiency, and maternal immigrant status (28, 29, 30, 31, 32, 33)? These factors and vaccines are simply unconnected.

Autism/ASD are clearly multi-factorial, with both genetic and environmental factors intersecting in as-yet undeciphered ways, and since rates started to increase dramatically since the 1980s, clearly some environmental factor(s) are key. However, those factors still remain stubbornly unclear. Rather than vaccines, however, multiple studies since at least 2004 have consistently reported altered gut microbiota composition in ASD subjects (10, 11). Whether that’s cause or effect still remains to be determined.


Basing anti-vaccine sentiment on a purported vaccines-autism link is reckless and dangerous since it inflicts real cost in the form of needless deaths from vaccine preventable diseases. Consider measles where the vaccine is historically one of the safest on record. In June 2017, a six year old Italian leukemia patient died from measles complications after reportedly catching it from his older brother, whom his parents had decided not to vaccinate (34), the latest in a measles ‘tragedy’ that has so far taken 35 lives across Europe (35).


1. Wakefield, Andrew J., et al. “RETRACTED: Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children.” (1998): 637-641. http://www.thelancet.com/pdfs/jo…

2. da Silva, Jaime A. Teixeira, and Judit Dobránszki. “Highly cited retracted papers.” Scientometrics 110.3 (2017): 1653-1661.

3. Collins, Harry M., Luis Reyes‐Galindo, and Paul Ginsparg. “A note concerning primary source knowledge.” Journal of the Association for Information Science and Technology 68.5 (2017): 1105-1110. https://arxiv.org/ftp/arxiv/pape…

4. Taylor, Brent, et al. “Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a causal association.” The Lancet 353.9169 (1999): 2026-2029. https://www.researchgate.net/pro…

5. Kaye, James A., Maria del Mar Melero-Montes, and Hershel Jick. “Mumps, measles, and rubella vaccine and the incidence of autism recorded by general practitioners: a time trend analysis.” Bmj 322.7284 (2001): 460-463. https://pdfs.semanticscholar.org…

6. Halsey, Neal A., and Susan L. Hyman. “Measles-mumps-rubella vaccine and autistic spectrum disorder: report from the New Challenges in Childhood Immunizations Conference convened in Oak Brook, Illinois, June 12–13, 2000.” Pediatrics 107.5 (2001): e84-e84. http://pediatrics.aappublication…

7. Demicheli, Vittorio, et al. “Vaccines for measles, mumps and rubella in children.” Cochrane Database Syst Rev 4.4 (2005). https://www.researchgate.net/pro…

8. Stratton, Kathleen, et al. “Immunization safety review: measles-mumps-rubella vaccine and autism.” (2001). https://www.ncbi.nlm.nih.gov/boo…

9. Evans, Bonnie. The metamorphosis of autism. Manchester University Press, 2017. https://www.ncbi.nlm.nih.gov/boo…

10. Mayer, Emeran A., David Padua, and Kirsten Tillisch. “Altered brain‐gut axis in autism: Comorbidity or causative mechanisms?.” Bioessays 36.10 (2014): 933-939.

11. Hsiao, Elaine Y. “Gastrointestinal issues in autism spectrum disorder.” Harvard review of psychiatry 22.2 (2014): 104-111. https://pdfs.semanticscholar.org…

12. nature.com search

13. nature.com search

14. nature.com search

15. Baron-Cohen, Simon, and Jessica Hammer. “Is autism an extreme form of the” male brain”?.” Advances in Infancy research 11 (1997): 193-218. https://pdfs.semanticscholar.org…

16. Whiteley, Paul, et al. “Gender ratios in autism, Asperger syndrome and autism spectrum disorder.” Autism Insights 2 (2010): 17. https://www.researchgate.net/pro…

17. Ruzich, Emily, et al. “Sex and STEM occupation predict autism-spectrum quotient (AQ) scores in half a million people.” PloS one 10.10 (2015): e0141229. Sex and STEM Occupation Predict Autism-Spectrum Quotient (AQ) Scores in Half a Million People

18. Baron-Cohen, Simon, et al. “Elevated fetal steroidogenic activity in autism.” Molecular psychiatry 20.3 (2015): 369. https://www.nature.com/mp/journa…

19. Muhle, Rebecca, Stephanie V. Trentacoste, and Isabelle Rapin. “The genetics of autism.” Pediatrics 113.5 (2004): e472-e486. https://www.researchgate.net/pro…

20. Rosenberg, Rebecca E., et al. “Characteristics and concordance of autism spectrum disorders among 277 twin pairs.” Archives of pediatrics & adolescent medicine 163.10 (2009): 907-914. https://www.researchgate.net/pro…

21. Hallmayer, Joachim, et al. “Genetic heritability and shared environmental factors among twin pairs with autism.” Archives of general psychiatry 68.11 (2011): 1095-1102. https://pdfs.semanticscholar.org…

22. Bailey, Anthony, et al. “Autism as a strongly genetic disorder: evidence from a British twin study.” Psychological medicine 25.1 (1995): 63-77. https://www.researchgate.net/pro…

23. Steffenburg, Suzanne, et al. “A twin study of autism in Denmark, Finland, Iceland, Norway and Sweden.” Journal of Child Psychology and Psychiatry 30.3 (1989): 405-416.

24. Folstein, Susan, and Michael Rutter. “A Twin Study of Individuals with Infantile Autism.” Autism. Springer US, 1978. 219-241.

25. Konstantareas, M. Mary, and Soula Homatidis. “Brief report: Ear infections in autistic and normal children.” Journal of autism and developmental disorders 17.4 (1987): 585-594.

26. Niehus, Rebecca, and Catherine Lord. “Early medical history of children with autism spectrum disorders.” Journal of Developmental & Behavioral Pediatrics 27.2 (2006): S120-S127.

27. Adams, James B., et al. “Mercury, lead, and zinc in baby teeth of children with autism versus controls.” Journal of Toxicology and Environmental Health, Part A 70.12 (2007): 1046-1051.

28. Landrigan, Philip J. “What causes autism? Exploring the environmental contribution.” Current opinion in pediatrics 22.2 (2010): 219-225. http://www.autism-society.org/wp…

29. Rossignol, Daniel A., and Richard E. Frye. “A review of research trends in physiological abnormalities in autism spectrum disorders: immune dysregulation, inflammation, oxidative stress, mitochondrial dysfunction and environmental toxicant exposures.” Molecular psychiatry 17.4 (2012): 389. https://pdfs.semanticscholar.org…

30. Grabrucker, Andreas M. “Environmental factors in autism.” Frontiers in Psychiatry 3 (2012). https://www.researchgate.net/pro…

31. Rossignol, D. A., S. J. Genuis, and R. E. Frye. “Environmental toxicants and autism spectrum disorders: a systematic review.” Translational psychiatry 4.2 (2014): e360. https://www.ncbi.nlm.nih.gov/pmc…

32. Ornoy, A., L. Weinstein-Fudim, and Z. Ergaz. “Prenatal factors associated with autism spectrum disorder (ASD).” Reproductive Toxicology 56 (2015): 155-169. https://www.researchgate.net/pro…

33. Ng, Michelle, et al. “Environmental factors associated with autism spectrum disorder: a scoping review for the years 2003-2013.” Chronic Diseases and Injuries in Canada 37.1 (2017). http://www.phac-aspc.gc.ca/publi…

34. Child’s death from measles caught from unvaccinated brother reignites debate in Italy

35. Measles ‘tragedy’ kills 35 across Europe – BBC News