Tool Street: Meta-analysis Never Sleeps

Following on from my post ‘Tool Street‘ where I was NOT mean about a stupid man spreading influential nonsense (thanks for your comment mum 🙂 ) I thought some people might find interesting this publication in the Journal of Evidence-Based Medicine. This meta-analysis from Li et al. has just been published online and concludes that:

‘oral sex is a marker rather than an independent risk factor of oral cancer because no significant association was detected between oral sex and oral cancer’.

So obviously more data and analyses needed to further reduce uncertainty. And obviously use condoms. But oral sex does not an oral cancer make.


Tool Street

In 2010 Michael Douglas announced that he had been diagnosed with stage IV throat cancer. In 2013 he gave an interview to The Guardian in which he stated that his throat cancer (later revised to tongue cancer) had been caused by the human papilloma virus (HPV) transmitted through oral sex. With his identification of HPV as the cause of his cancer he also said that he believed that his other lifestyle choices – smoking and drinking – were not responsible. He also asserted that oral sex was the best cure for his cancer.

Last year, Bruce Dickinson was diagnosed with stage III tongue cancer, and he announced last week that his doctors have identified HPV as the cause of his cancer. He managed not to deny that that drinking, smoking and drug-taking could be contributors to his cancer, or promote oral sex are a cure.

Bruce 1, Michael 0.

Michael Douglas has since said he regrets the comments he made – more specifically – he regrets the embarrassment that they caused his wife and her family. No mention of the dangerous medical myths he has promoted to a huge audience, the idea that HPV is the cause of tongue cancer, other lifestyle factors aren’t, and that sex can cure cancer. It also completely overlooks cervical cancer, which is much more often caused by HPV.

Where does that leave us? With two famous men, respected in their professional fields with many fans, and extremely well publicised illnesses, both caused (apparently solely) by HPV. It’s hard to see how any correction of the inaccuracies in Michael Douglas’s story, or accurate information about HPV can ever reach the same audience.

Electron micrograph of HPV

Electron micrograph of HPV

The main causes of oral cancers are smoking (linked to 65% of cases), drinking (linked to 30% of cases), diet (insufficient diet is linked to 56% of cases), and HPV. Age and genetics also affect risk. To ignore the huge effect of these causes other than HPV is irresponsible. Men are twice as likely as women to get tongue cancer, they are also more likely to smoke more, drink more, and eat less well. With evidence it’s easy to refute these myths.

Most people have an HPV infection. Relatively few people get oral cancer. Not all types of HPV cause cancer, and not all cancer-causing HPV infections cause oral cancers. In fact, HPV infections cause more cases of cervical cancer than oral cancers. Oral cancers are more common overall, but HPV is not the main cause of them. It is the main cause of cervical cancers.

In order to reduce the risk of tongue cancer people need to know which activities are higher risk (smoking, drinking, poor diet) and which are lower risk (sex), so they can choose which to do, or not do. In order to prevent cervical cancer, preventing HPV infection is key. There is an effective and safe HPV vaccine. It is estimated that in the UK it will halve the number of deaths from cervical cancer each year.


Tongue cancer

HPV is a cause but not one of the main causes.

Smoking, drinking and poor diet are the main causes.

Cervical cancer

HPV infection is the main cause and can be prevented.

Michael Douglas

Is a tool.

Don’t Drink the Water

Actually, most of you reading this – do. It’s almost certainly not going to hurt you.

Affected area of Lancashire - boil your water.

Affected area of Lancashire – boil your water.

In parts of Lancashire in the UK the current advice is to boil your water, due to a very low level of Cryptosporidium detected at Franklaw water treatment works in Preston. This advice should be followed – why risk getting Cryptosporidiosis (diarrhoea, stomach pains, cramps and low fever) when you don’t have to? Although the infection is generally not serious in healthy people, in some, and in those who are immunocompromised (cancer patients on chemotherapy, babies and young children, older people) a more severe form of the disease can develop (with symptoms including chronic diarrhoea).

Cryptosporidium parvum oocysts from the EPA

Cryptosporidium parvum oocysts (from the EPA)

It’s a faff to boil water before using it, and surely some people are alarmed and upset by the risk of infection, but relatively, we’re lucky. Low levels of one genus of parasite, one water treatment works contaminated temporarily, the means to kill the parasite, and free care in the event that infections do occur. But, parasites in our drinking water – it’s a surprise to get a small taste of the way most people have to live.

Entamoeba histolytica trophozoite by Iqbal Osman

Entamoeba histolytica trophozoite by Iqbal Osman

There are many parasites and genera, including Cryptosporidium, that cause disease by infecting people through their water supply: Entamoeba histolytica, Dracunculus medinensis, Schistosoma, Giardia duodenalis, Cyclospora cayetanensis, Toxoplasma gondii, Cystoisospora belli, Blastocystis hominis… the list goes on. And these are just (just!) some of the parasites – there are plenty of bacteria and viruses too…

Dracunculus medinensis by CDC

Dracunculus medinensis by CDC

The World Health Organisation estimates that 842,000 deaths per year are attributable to unsafe drinking water, sanitation and hygiene. 261,000 of those deaths are of children under the age of five.

This infographic from the Center for Disease Control shows all the countries where they recommend that we do not drink the tap water – most countries, and where most people live. The Millenium Development Goals drinking water target has been met – now 89% of the world’s people have access to ‘improved water sources’, but the remaining 748,000,000 people don’t. People in rural areas and those who are poor are less likely to have access to safer water.

Schistosoma by David Williams, Illinois State University

Schistosoma by David Williams, Illinois State University

The world-wide battle to kill parasites in drinking water continues. Most parasites can be killed by boiling drinking water, but many people do not have the resources to boil water. Filtering the water removes many parasites, but effective filters are expensive, and make-shift filters are less effective. Addition of chemicals, such as chloride, kills many organisms, but not others. UV irradiation of water can kill organisms, but doesn’t work with cloudy water, few people have access to such facilities, and UV irradiation using sunlight is less reliably effective. For many people, infection by parasites is inevitable; people in rural areas and those who are poor are less likely to be able to make their water safer.

Humans fibroblasts infected with Toxoplasma gondii by Maria Francia & Boris Striepen, University of Georgia

Humans fibroblasts infected with Toxoplasma gondii by Maria Francia & Boris Striepen, University of Georgia

Many people do not have the resources to be diagnosed with an infection. If they are diagnosed, many cannot be treated. Cryptosporidosis in severe cases is treated by supportive care (rehydration). Supportive care is not available to many people; severe dehydration can lead to death. Amoebiasis can be treated with amaoebicides but many people don’t have access to them. There is no treatment for Dracunculiasis (guinea worm disease), and removal of the worm is difficult. Schistosomiasis can be treated, but many people don’t have access to medication… You get the idea.

Parasitology is an under-researched subject and an under-funded medical field. Much of the funding and research there is goes on malaria (as much of it should), but that leaves other diseases neglected and we just aren’t spending enough money on dealing with these preventable and eradicable diseases. There are programs for prevention, treatment and eradication of some of these parasitic infections, these programs will also reduce bacterial and viral infections. Cleaner water, improved sanitation, access to medical care and more research are all essential if we want to save lives.

Don’t Drink the Water

31/08/2015 Boil water notice lifted in some parts of Lancashire

21 Days Later

The vaccine is here! The one we dared to dream of! 100% success!

Well… not exactly.

rVSV-ZEBOV is a promising Ebola virus vaccine candidate.

A trial of this vaccine candidate was being carried out in Basse-Guinée, and an interim analysis of this work was published on the 31st July in The Lancet. The data has generated much excitement and enthusiasm, and the coverage has left me with a feeling that perhaps not everyone who should have read the paper, did read the paper…

Ebola vaccine trial proves 100% successful in Guinea

Picture 3

The Ebola vaccine we dared to dream of is here

It looks like a done deal from these stories, but there has been a much more measured reaction to the data by the authors of the paper, those closely involved in the trial, and other media. Caveats abound, as they should with such early and limited data.

The WHO Assistant Director, Marie-Paul Kieny spoke at a press conference: “The data so far – and it’s as I said, an intermediary analysis and the trial is going on – but the data so far, shows that none of the 2,014 persons vaccinated developed Ebola virus disease after 10 days after vaccination.”

The paper itself states: ‘The results of this interim analysis indicate that rVSV-ZEBOV might be highly efficacious and safe in preventing Ebola virus disease, and is most likely effective at the population level when delivered during an Ebola virus disease outbreak via a ring vaccination strategy.’

I’m thrilled that we do have positive results. I work at St. George’s, where I worked on, and where work continues on, a Phase I clinical trial of this vaccine. I want my vaccine to work. I’m not a sociopath; I want any vaccine to work. But I am a skeptic. I want the vaccine to actually work, not just to look like it works, and I want to understand how well it works, and how it can be used effectively and safely.

This is the first published data from a phase III clinical trial of an Ebola virus vaccine candidate. Clinical trials perform the purpose of reducing our uncertainty about an intervention. The more high-quality trials we have, the less uncertain we can be about how well an intervention works. This is one trial. It has reduced our uncertainty. These are interim results – some of the results, from some of the trial – they are not the results of the full trial. There are more data, and more types of data in this trial, and more trials that we need to look at before we can be less uncertain about how well the vaccine works, and how safe the vaccine is.

If it is to be THE Ebola virus vaccine, we also need to work out who should be vaccinated, who shouldn’t, and how this new vaccination program could be managed. Although the results are very hopeful, it is a bit more complicated than 100% success achieved.

Other good news will make vaccine research even more difficult – we will (hopefully) have fewer and fewer patients on whom to test the vaccine. The number of new cases of Ebola virus disease in the week to 26th July was seven – down from the 20-30 cases per week it had been for the two previous months.

The trial

It is a great trial design. This was an open-label, cluster-randomised ring vaccination trial. For each new confirmed case, all the people who had been in close contact with that person, and all the people who had been in contact with them were defined as a cluster. Each cluster was randomly allocated to receive the vaccine immediately or after 21 days. The number of cases of Ebola virus disease in the two groups was then compared.

The results

The figure of 100% efficacy was arrived at by comparing the number of people in the group who were vaccinated immediately and got Ebola virus disease after 10 or more days, with the number of people in the group who were vaccinated after 21 days and got Ebola virus disease after 10 or more days. Data for 48 clusters in which 2014 patients were vaccinated immediately, and 42 clusters in which 1498 patients were vaccinated 21 days later were used in this analysis. There were no cases of Ebola virus disease after 10 days in the immediate vaccination group, whereas in the after 21 days vaccination group there were 16 cases of Ebola virus disease, so vaccine efficacy of 100% (p=0·0036).


What about the people who got Ebola virus disease within 10 days after vaccination? 9 of the 2014 immediately vaccinated patients did get Ebola virus disease within 10 days, compared to 16 of the 1498 patients who were not immediately vaccinated. So this study found that the vaccine is not (and of course it was not expected to be) 100% effective in preventing Ebola virus disease from days 0-21.

And that isn’t the only data, or the only data we need.

The study did look at adverse events, and mentions 43 serious adverse events – but analysis is not complete and details were not published in this paper. We need to know more. Side effects of the vaccine might be severe enough and common enough to outweigh the possible benefits of the vaccine, especially in areas where people are at an extremely low risk of infection. Long-term sequelae have not yet been studied.

What about other people? This study did include not children, or some women – we need to look at outcomes in these groups.

What about other doses? Might a smaller dose be equally effective, but have fewer side effects? Phase I preliminary results (full results not yet published) showed a number of adverse events in (unexposed) patients receiving lower doses of the vaccine than was used here, but were they less common and/or less severe?

An electron micrograph of an Ebola viral particle By CDC/ Dr. Frederick A. Murphy

An electron micrograph of an Ebola viral particle
By CDC/ Dr. Frederick A. Murphy

Whom will this vaccine benefit? How many people will have to be vaccinated to prevent one infection? How many to prevent an outbreak? What method of vaccination? How long would any protection last? With our limited pot of money for dealing with health issues, how much should be spent on vaccination against Ebola virus disease, when other preventable diseases kill so many more people?

Due to the positive results of this trial, it has been agreed that it is no longer ethical to have a group of people who are vaccinated after 21 days; instead everyone will be vaccinated immediately. This leaves us without a control group, which makes it difficult to determine how many infections have been prevented. Control groups are also required to study possible side effects. We’re without many patients to study, as the outbreak hopefully comes to an end.

These data is a huge achievement, many thousands of people have worked together to get to this point, and it is very encouraging. BUT, we need more data to further reduce our uncertainty about the efficacy and safety of the vaccine. We don’t know what the outcomes of this trial or other ongoing trials of this vaccine will be. We can now be more hopeful that they will be positive.

We don’t have an Ebola virus vaccine yet, but we do have a promising Ebola virus vaccine candidate.

There Will Be Bees

Bees. People love bees. Unless they have an allergy to bee stings. Or a phobia of bees. Or a picnic, or an ice-cream… Bloody bees.

Adorable. Photograph by André Karwath

Photograph by André Karwath

Maybe it would be more accurate to state that bees make great headlines right now.

Guess what don’t? Parasites.

And so, exploiting the popularity of bees to push my own agenda [insert parasite joke], and in the hope of dragging in readers who aren’t in love with infectious diseases, have a look at this.

Bees. Without them we would lose much of our pollination; we’d have lower yields from food crops (about one quarter of our crops rely on pollination by honey bees), a reduction in plant diversity, and a reduction in the animal species that rely on them. That includes us.

The European honey bee, Apis mellifera, is one of the world’s most important pollinators. Nosema apis is a microsporidian, a parasite that infects the gut of A. mellifera. N. apis causes nosemosis – dysentery, crawling (rather than flying) increased girth of abdomen, missing sting reflex and early replacement of the queen bee. It has been assumed to be of low virulence, but might an N. apis infection, as well as causing early death, be having an effect on the ability of surviving bees to pollinate?

The decline in the number of bees has been the main consideration in reduced pollination, but changes in the behaviour in surviving pollinators may also significantly negatively affect pollination. This is the first paper to track foraging behaviour of individual N. apis infected bees.

“Parasitized honey bees are less likely to forage and carry less pollen.”

This research looked at 960 A. mellifera worker bees and gave them either sucrose solution or N. apis spores in sucrose solution, and tagged the bees with unique radio frequency identification tags (awesome) to monitor their foraging behaviour.


A tagged bee! Photograph by Lori Lach

A tagged bee!
Photograph by Lori Lach

  • Spore-fed bees were less likely to forage than sucrose-only fed bees (p =0.0029)
  • Spore-fed bees that did forage started older (p=0.04) and stopped younger (p=0.008).


  • There was not a significant difference between the groups in:
    • the number of foraging trips taken per day (p=0.08).
    • the total hours foraged over a lifetime (p=0.19).
    • homing ability.
Another tagged bee! Photograph by Lori lach

Another tagged bee!
Photograph by Lori lach


  • Spore-fed bees were 4.3 times less likely to be carrying available pollen.
  • The number of pollen grains carried was negatively correlated with the number of apis spores (p=0.009).
  • There was a difference (p=0.061) in the choice of spore-fed bees to choose artificial sugar flowers over artificial pollen flowers, while sucrose-fed bees visited both equally.

The results are mixed. In this paper, N. apis infection of bees affected some measures of pollination behaviour and pollination carriage, and not others. Innoculated bees were more likely to die earlier, as had been found in previous studies. Precocious pollination had been found in previous studies, but was not seen here.

N.apis infection may have a negative effect on the pollination behaviour and carriage of A. mellifera – bee parasites might be reducing bee pollination.


And how did the researchers attach the unique radio frequency identification tags to the bees to measure all this?

Even more tagged bees! Photograph by Lori Lach

Even more tagged bees! Photograph by Lori Lach

We just had to hold them in our hands and hope the glue dried quickly. It was actually quite a process – they had to be individually painted, then individually fed, then the tag glued on. Then individually scanned so we knew which tag was on what color and treatment bee and which hive it was going into. It all had to happen within about eight hours of emergence because as the day goes on they start learning how to fly and they get better at stinging.

Dr Lori Lach, James Cook University


Edited 15/07/15 to correct spelling (Mellifera), 16/07/15 to correct mistake (the first paper to track foraging behaviour of individual N. apis infected bees.)

20,000 leagues under the MERS

This week it was confirmed that there has been a second case of MERS in the Philippines.

Middle East Respiratory Syndrome and the virus causing it, MERS coronavirus, were first identified in Saudi Arabia in 2012, and since then cases have been popping up across the globe; Iran, Jordan, Kuwait, Lebanon, Oman, Qatar, Saudi Arabia, the United Arab Emirates, and Yemen (Middle East); Austria, France, Germany, Greece, Italy, Netherlands, Turkey, and the United Kingdom (Europe); Algeria, Tunisia and Egypt (Africa); China, Malaysia, Republic of Korea, the Philippines and Thailand (Asia); and the United States of America (Americas) all have laboratory-confirmed cases. Since September 2012 there have been 1,365 cases and 487 MERS-related deaths. We still aren’t sure of the reservoir hosts, or the route of transmission from these hosts, or how easily human-to-human transmission occurs. There’s no vaccine and no treatment (other than supportive).

Transmission electron micrograph of Middle Eastern Respiratory Syndrome CoV particles found near the periphery of an infected MRC-5 cell. Credit: NIAID

Transmission electron micrograph of Middle Eastern Respiratory Syndrome CoV particles found near the periphery of an infected MRC-5 cell. Credit: NIAID

Coverage of MERS by the UK media has ranged from trivial to distractingly Eurocentric, numerically irresponsible, fear-mongering, and hilarious (although terrifying in a completely different way).

So how worried do we need to be? How does MERS-CoV compare to other viral infectious diseases?

Fewer confirmed cases: SARS is another coronavirus that causes a respiratory syndrome; in less than a year, SARS caused approximately 8,096 cases and 774 deaths, and has now been eradicated.

Fewer confirmed deaths: In 2012 HIV/AIDS caused approximately 1.5 million deaths, 3000 times more than MERS-CoV in three years.

Less infectious: Measles virus is highly contagious (R0 = 12-18), infecting 90% of non-immune contacts, whereas few infections of MERS-CoV have been passed between humans without close contact (R0 = 2-6.7).

Manageable:  Other viral outbreaks such as Ebola have been tackled with some success; the spread of the virus is now more controlled and numbers of cases are dropping. We have developed rapid diagnostic tests, specialist testing laboratories, a global network of experts and health professionals and made great progress in vaccine development.

Colorized scanning electron micrograph of Middle Eastern Respiratory Syndrome virus particles attached to the surface of an infected VERO E6 cell. Credit NIAID

Colorized scanning electron micrograph of Middle Eastern Respiratory Syndrome virus particles attached to the surface of an infected VERO E6 cell. Credit NIAID

However, the numbers of cases and deaths do not take into account the number of deaths that are, but have not been discovered to be MERS-related. They don’t take into account the number of cases that were symptomatic but not discovered to be MERS-related. And they don’t take into account the number of asymptomatic infections. There may be many more cases of, and deaths due to, MERS-CoV than we know about.

We still don’t know enough about how MERS-CoV is transmitted. It seems fairly certain that it is transmitted through from person-to-person via the infected person’s respiratory secretions, but we don’t know the role of other reservoirs, including bats and camels, and whether direct or indirect contract is required for transmission.

With so many gaps in our understanding of MERS it will be difficult to reduce transmission and develop resources to tackle the outbreak. With so few cases it will be difficult to study the virus or test any potential treatments or vaccines. And with cases in so many countries it will be difficult to coordinate these efforts and eradicate MERS-CoV. Viruses and outbreaks are unpredictable and can change, it will be difficult to anticpate this.

Transmission electron micrograph of Middle East respiratory syndrome coronavirus. Credit: NIAID

Transmission electron micrograph of Middle East respiratory syndrome coronavirus. Credit: NIAID

But, there’s plenty of work ongoing; although we don’t know everything we need to or have perfect health systems in place, we are making progress. We know who the virus is likely to infect, what areas it is more common in, and the symptoms that the virus is likely to cause. We can diagnose previous exposure using ELISA, IFA and microneutralisation, and current infection using rRT-PCR. We know that supportive treatment helps. Whole genome sequences have been published. We know that it may have originated in bats and passed to camels, and that camel-to-human and human-to-human transmission are possible. It has been possible to make sensible recommendations to reduce the risk of infection and prevent spread of the disease.

Perhaps it’s not something to cause panic, but something to fund research into, to understand, treat, prevent and eradicate.

Dawn of the Spread (of dengue virus vectors)

A paper published today in the journal eLIFE describes work to model the distribution of two species of mosquito, Aedes aegypti (yellow fever mosquito) and Aedes albopictus (Asian tiger mosquito). These two species are vectors of dengue, yellow fever, and chikungunya viruses; their presence may precede autochthonous cases of these diseases. The model developed by these researchers is based on recorded occurrences of mosquitoes, their eggs or larvae (the largest number of records of this type ever compiled) and environmental factors including temperature and type of land cover.

Aedes aegypti” by James Gathany – CDC

The model predicts that these two vectors are spreading throughout the world, and are present on all continents. The authors hypothesise  that where the vectors are, the virus and disease might follow.

Existing research has already documented that cases of dengue fever are appearing in new geographic regions. With an increasing geographical spread of vector and virus we can expect an increase in the incidence and disease burden of dengue fever. Dengue fever is one of the better funded neglected tropical diseases, but there is not yet any treatment or vaccine licensed for use. Treatment is supportive only, and consists of rehydration and sometimes blood transfusion. There is hope of a vaccine, one tested in phase III clinical trials in Asia and Latin America in 2014 has shown mixed results.

However, the threat of spreading vectors and increasing cases of dengue fever in Europe are not the only reasons we might be interested in tackling this disease.

Currently, dengue virus causes 100 million apparent infections each year and is a leading cause of serious illness and death in children in some Asian and Latin American countries. Common symptoms include fever, headache, muscle and joint pain, and skin rashes, and the disease can progress to dengue haemorrhagic fever, with bleeding, low platelets, plasma leakage and low blood pressure. Within the last few months outbreaks and high numbers of cases of dengue fever have been reported from many countries including; Yemen, Kuala Lumpur, India, Cambodia, Malaysia, Thailand and Brazil. In countries where a stable and comprehensive health system is not in place, it will be very difficult or impossible to contain outbreaks without international help.

Aedes albopictus” by James Gathany, CDC

This research has mapped the global distribution of dengue virus vector, which is essential for public health planning. If we know where and in what numbers we might encounter cases of dengue fever we are more likely to succeed in preventing and containing outbreaks. Long-term investment in health systems and research will be key in controlling, preventing, treating, and eventually eradicating this disease.