The switch from trivalent to bivalent oral poliovirus vaccine: Will it lead to polio?

bivalent OPVIn four months, 155 countries will together switch from using trivalent to bivalent oral poliovirus vaccine. Will this change lead to more cases of poliomyelitis?

There are three serotypes of poliovirus, each of which can cause paralytic poliomyelitis. The Sabin oral poliovirus vaccine (OPV), which has been used globally by WHO in the eradication effort, is a trivalent vaccine that contains all three serotypes.

In September 2015 WHO declared that wild poliovirus type 2 has been eradicated from the planet – no cases caused by this serotype had been detected since November 1999. However, in 2015, there were 9 cases of poliomyelitis caused by the type 2 vaccine. For these reasons WHO decided to remove the type 2 Sabin strain from OPV, and switch from trivalent to bivalent vaccine in April 2016.

After OPV is ingested, the viruses replicate in the intestinal tract, providing immunity to subsequent infection. During replication in the intestine, the vaccine viruses lose the mutations that prevent them from causing paralysis. Everyone who receives OPV sheds these revertant viruses in the feces. In rare cases (about one in 1.5 million) the revertant viruses cause poliomyelitis in the vaccine recipient (these cases are called VAPP for vaccine-associated paralytic poliomyelitis). Vaccine-derived polioviruses can also circulate in the human population, and in under-vaccinated populations, they can cause poliomyelitis.

There were 26 reported cases of poliomyelitis caused by the type 1 or type 2 vaccine viruses in 2015. Nine cases of type 2 vaccine-associated polio were detected in four countries: Pakistan, Guinea, Lao People’s Democratic Republic, and Myanmar. Removing the type 2 strain from OPV will eliminate vaccine-associated poliomyelitis in recipients caused by this serotype. When the US switched from OPV to the inactivated poliovaccine (IPV) in 2000, VAPP was eliminated.

The problem with the trivalent to bivalent switch is that vaccine-derived type 2 poliovirus is likely still circulating somewhere on Earth. The last two reported cases of type 2 vaccine-associated polio in 2015 were reported in Myanmar in October. The viruses isolated from these cases were genetically related to strains that had been circulating in the same village in April of the that year. In other words, type 2 vaccine-derived strains have been circulating for an extended period of time in Myanmar; they have been known to persist for years elsewhere. If these viruses continue to circulate past the time that immunization against type 2 virus stops, they could pose a threat to the growing numbers of infants and children who have not been immunized against this serotype.

Eventually as type 3, and then type 1 polioviruses are eradicated, it will also be necessary to stop immunizing with the respective Sabin vaccine strains. The switch from trivalent to bivalent vaccine in April 2016 is essentially an experiment to determine if it is possible to stop immunizing with OPV without placing newborns at risk from circulating vaccine-derived strains.

Over 18 years ago Alan Dove and I argued that the presence of circulating vaccine-derived polioviruses made stopping immunization with OPV a bad idea. We suggested instead a switch from OPV to IPV until circulating vaccine-derived viruses disappeared. At the time, WHO disagreeed, but now they recommend that all countries deliver at least one dose of IPV as part of their immunization program. Instead of simply removing the Sabin type 2 strain from the immunization programs of 155 countries, it should be replaced with the inactivated type 2 vaccine. This change would maintain immunity to this virus in children born after April 2016. Such a synchronized replacement is currently not in the WHO’s polio eradication plans. I hope that their strategy is the right one.

TWiV 371: Sympathy for the devil

TWiVOn episode #371 of the science show This Week in Virology, the TWiVologists discuss the finding of a second transmissible cancer in Tasmanian devils, and development of new poliovirus strains for the production of inactivated vaccine in the post-eradication era.

You can find TWiV #371 at

Virologists, start your poliovirus destruction!

I have worked on poliovirus for over thirty-six years, first as a posdoctoral fellow with David Baltimore in 1979, and then in my laboratory at Columbia University. The end of that research commences this year with the destruction of my stocks of polioviruses.

In 2015 there were 70 reported cases of poliomyelitis caused by wild type 1 poliovirus, and 26 cases of poliomyelitis caused by circulating vaccine derived polioviruses (cVDPV) types 1 and 2. The last case of type 2 poliovirus occurred in India in 1999, and the virus was declared eradicated in 2015. Consequently the World Health Organization has decided that all remaining stocks of wild type 2 poliovirus should be destroyed by the end of 2015.

My laboratory has worked extensively with type 2 polioviruses. Before we produced transgenic mice susceptible to poliovirus, we had studied the Lansing strain of type 2 poliovirus because it had the unusual ability to infect wild type mice (polioviruses normally only infect certain primates). We determined the nucleotide sequence of the viral genome, identified the capsid as a determinant of the ability of the virus to infect wild type mice, and showed that swapping an eight amino acid sequence of capsid protein VP1 from a type 1 strain with that from Lansing conferred the ability to infect non-transgenic mice. These findings indicate that the ability of the Lansing strain of poliovirus to infect mice is likely due to recognition by the viral capsid of a receptor in the mouse central nervous system. In the past year we took advantage of the ability to produce mouse neurons from stem cells to attempt to identify the murine cellular receptor for Lansing virus.

To prevent further cases of poliomyelitis caused by cVDPVs, WHO has decided that there will be a synchronized, global switch from trivalent OPV to bivalent OPV in April 2016. By July of 2016 all remaining stocks of the Sabin type 2 poliovirus strains, which are used to produce OPV, will also be destroyed.

No wild type 3 poliovirus has been detected since November 2012, and it is likely that this virus will be declared eradicated within the next several years. At that time we will have to destroy our stocks of type 3 poliovirus. That leaves wild poliovirus type 1, which circulates only in Pakistan and Afghanistan. Given the small number of cases of paralysis caused by this type, it is reasonable to believe that eradication will occur within the next five years. If this timeline is correct, it means that I will be destroying my last vials of poliovirus around 2020.

It is of course necessary to destroy stocks of wild and vaccine polioviruses to prevent reintroduction of the virus and the disease that it causes. The 1978 release of smallpox virus from a laboratory in the United Kingdom, which caused one death, lead to requests for reducing the number of laboratories that retained the virus. Today there are just two official repositories of smallpox virus in the United States and Russia.

It is rare for an investigator to be told to destroy stocks of the virus that is the subject of his or her research. Over the years we have published 81 papers on poliovirus replication, vaccines, and pathogenesis. While I realize that it is absolutely essential to stop working on this virus, I do so with a certain amount of sadness. What other emotion could I have for a virus on which I have expended so much thought and effort?

Image: Poliovirus by Jason Roberts

Correction: The synchronized switch in April 2016 is from trivalent to bivalent OPV, not OPV to IPV. Consequently I have removed comments related to an OPV-IPV switch.

TWiV 367: Two sides to a Coyne

On episode #367 of the science show This Week in Virology, two Coynes join the TWiV overlords to explain their three-dimensional cell culture model of polarized intestinal for studying enterovirus infection.

You can find TWiV #367 at

TWiV 353: STING and the antiviral police

On episode #353 of the science show This Week in Virology, the TWiVniacs discuss twenty-eight years of poliovirus shedding by an immunodeficient patient, and packaging of the innate cytoplasmic signaling molecule cyclic GMP-AMP in virus particles.

You can find TWiV #353 at

Shedding poliovirus for 28 years

Glass PoliovirusAn immunodeficient individual has been excreting poliovirus in his stool for 28 years. Such chronic excreters pose a threat to the poliovirus eradication program.

Since its inception in 1988 by the World Health Organization, the poliovirus eradication program has relied on the use of the infectious, attenuated vaccine strains produced by Albert Sabin. These viruses are taken orally, replicate in the intestine, and induce protective immunity. During replication in the gut, the Sabin strains lose the mutations that prevent them from causing paralysis. Nearly every individual who receives the Sabin vaccine strains excretes so-called vaccine-derived polioviruses (VDPVs) which are known to have caused outbreaks of poliomyelitis in under-immunized populations.

Immunocompromised individuals who produce very low levels of antibodies (a condition called agammaglobulinemia) are known to excrete VDPVs for very long periods of time – years, compared with months in healthy individuals. Seventy-three such cases have been described since 1962. These individuals receive the Sabin vaccine in the first year of life, before they are known to have an immunodeficiency, at which time they must receive antibodies to prevent them from acquiring fatal infections.

The most recently described immunocompromised patient was found to excrete poliovirus type 2 vaccine for 28 years (the time period is determined by combining the known rate of change in the poliovirus genome with sequence data on viruses obtained from the patient).  The VDPV is neurovirulent (causes paralysis in a mouse model), antigenically drifted, and excreted in the stool at high levels.

Because the polio eradication plan calls for cessation of vaccination at some future time, these immunocompromised poliovirus shedders pose a threat to future unimmunized individuals. The global number of such patients is unknown, and there is no available therapy to treat them – administration of antibodies does not clear the infection. The development of antivirals that could eliminate the chronic poliovirus infection is clearly needed (and ongoing). It will also be necessary to conduct environmental surveillance for the presence of VDPVs – they can be identified by properties that distinguish them from VDPVs produced by immunocompetent vaccine recipients.

While the WHO eradication plan now includes a shift to using inactivated (Salk) poliovaccine, this strategy would not impact the existing immunocompromised poliovirus shedders. Should a VDPV from these individuals cause an outbreak of polio in the post-vaccine era, it will be necessary to control the outbreak with Salk vaccine, or an infectious poliovirus vaccine that cannot revert to virulence during replication in the intestine. Polioviruses with a recoded genome are candidates for the latter type of vaccine.

Image credit: Jason Roberts

A collection of polioviruses

PoliovirusesIn midsummer 1986, five years after starting my poliovirus laboratory at Columbia University, I received a letter from Frederick L. Schaffer, a virologist at the University of California, Berkeley, asking if I would like to have his collection of poliovirus stocks. He was retiring and the samples needed a home, otherwise they would be destroyed. Of course I jumped at the opportunity to have a bit of virology history.

Three boxes full of dry ice arrived in the laboratory in August 1986. In them were sixty-six containers of different polioviruses that Dr. Schaffer had collected over the years. All three poliovirus serotypes were represented, both wild type and vaccine strains. The tubes were marked with dates ranging from 1948 to 1965. Most had come into Dr. Schaffer’s hands from well known poliovirologists, including Jonas Salk, Albert Sabin, Igor Tamm, Renato Dulbecco, and Charles Armstrong.

All of the samples were in glass containers, either tubes with a screw cap or a rubber stopper held in place with tape (pictured). A few were in glass bottles of the type that were used to grow cells. The collection held not a single plastic tube: these were the days before plastics entered the virology world. All were identified by hand-written labels on white cloth tape. Some of the labels in the photo read: Polio 1 Brunhilde (1963), Polio 2 MEF1, Polio 2 P712, HeLa P3, 10-21-62. Nearly all the samples were virus-containing cell culture supernatants.

Lansing poliovirusPerhaps the most amazing sample was a specimen labeled ‘Lansing poliomyelitis virus, 9/27/50, passage 379, C. Armstrong, NIH’. Within the glass vial was a intact mouse brain still attached to the spinal cord (there were no cell phones in 1986, hence no photo). The Lansing type 2 strain of poliovirus had been adapted to grow in mice by Charles Armstrong, and this was apparently the 379th intracerebral mouse-to-mouse passage! It was especially exciting for me to receive this sample, because my laboratory had been studying the ability of the Lansing strain of poliovirus to infect mice. I believe that the note accompanying the tube is in Dr. Armstrong’s writing.

I have since transferred most of the samples to plastic tubes which are stored in a -70C freezer. There is a trove of information to be obtained by studying these samples, but there are few poliovirologists left who are interested. Once poliomyelitis is eradicated – perhaps within the next 10 years – these samples, and similar ones throughout the world, will have to be destroyed.

The Wall of Polio, version 3.0

Wall of Polio 3.0Back in 2013 I built a Wall of Polio in my laboratory – a large stack of six-well cell culture plates that have been used to measure the concentration of polioviruses in various samples by plaque assay. It became a focal point of the lab at which many guests came to have their photographs taken. Sadly, the Wall fell twice. Now a new Wall – version 3.0 – has been completed.

The new Wall of Polio is in my office at Columbia University Medical Center, where it will not annoy the Fire Inspector (the former Wall partially blocked an aisle). Furthemore, the new Wall is glued together, so it will not come apart. Its construction is documented in the photographs below. The Wall of Polio 3.0 is built with 1,464 six-well plates of HeLa cells that were used to determine the titer of poliovirus. We have also already had a number of visitors to Wall 3.0.

Because the Wall is impressive, it attracts attention, which can then be used to explain the plaque assay and determining virus titer. Therefore it is simply another tool that I used to teach the world about virology.

When you visit, expect that I will ask to photograph you before the Wall. Only a few have refused.


TWiV 331: Why is this outbreak different from all other outbreaks?

On episode #331 of the science show This Week in Virology, the TWiV team discusses the possible association of the respiratory pathogen enterovirus D68 with neurological disease.

You can find TWiV #331 at

Covering up a naked virus

Sabin type 2 poliovirusViruses can be broadly classified according to whether or not the particle is enveloped – surrounded by a membrane taken from the host cell – or naked. Some naked viruses apparently are more modest than we believed.

Members of the family Picornaviridae, which include Hepatitis A virus, poliovirus, and Coxsackieviruses, have non-enveloped particles that consist of a protein shell surrounding the viral RNA genome (poliovirus is illustrated). Examples of viruses that are enveloped include dengue virus, influenza virus, and measles virus.

Recently it was discovered that hepatitis A virus (HAV) particles are released from cells in membrane vesicles containing 1-4 virus particles. These membranous structures resemble exosomes, which are also released from uninfected cells and play roles in various biological processes. Enveloped hepatitis A virus particles are present in the blood of infected humans. However virus in the feces, which is transmitted to other hosts, is not enveloped.

Viral envelopes typically contain viral glycoproteins, such as the HA protein of influenza viruses, which serve important functions during replication, such as attachment to cell receptors. Envelope glycoproteins are also the target of antibodies that block viral infection. The presence of an envelope makes HAV resistant to neutralization with antibodies, because the membrane contains no viral proteins that can be blocked by antibodies.

Two other non-enveloped picornaviruses, Coxsackievirus B and poliovirus, are also released from cells within membrane vesicles. These virus particles are in vesicles derived from the autophagy pathway, which captures and recycles cytoplasmic contents by ejecting them from the cell.

What is the function of the membrane acquired by these naked viruses? Perhaps immune evasion: the presence of the cell membrane makes HAV and Coxsackievirus B virus particles resistant to neutralization with antibody. The ability to deliver multiple virus particles to a single cell might help to overcome genetic defects in the viral genome that are a consequence of the high mutation rates of these viruses.

An interesting problem is how these cloaked viruses enter cells, because there is no evidence that the membranes contain any viral proteins that could interact with a cell receptor. Nevertheless, entry of enveloped HAV and poliovirus into cells requires the known viral receptor. Perhaps the vesicles are taken into the cell by endocytosis, where viral particles are released from the vesicles, and then bind receptors to initiate escape of the genome.

Should HAV, poliovirus, and Coxsackievirus B be reclassified as enveloped viruses? Probably not, in part because the membranes surrounding these virus particles are not needed for infectivity. In contrast, removal of the membrane from influenza virus, dengue virus, or measles virus destroys their infectivity. Enveloped viruses acquire a membrane after the internal components have been assembled, whether they are helical or icosahedral nucleocapsids. In contrast, HAV, poliovirus, and Coxsackievirus B become fully infectious particles before they acquire an envelope.

Another argument against calling picornaviruses enveloped is that viral membranes contain viral glycoproteins that allow attachment to cell receptors and release of the viral genome into the cell. There is no evidence that the membranes of picornaviruses contain viral proteins.

The acquisition of a membrane may have taken place later in the evolution of picornaviruses, to allow more efficient infection or evasion of host responses. Alternatively, the membrane may simply be a by-product acquired when these viruses exit the cell by a non-lytic mechanism.

While the finding of membranes around picornavirus particles is intriguing, I am not yet convinced that these viruses should be considered to be enveloped. I would like to know if other non-enveloped viruses are similarly released from cells in membranous cloaks, and the function of this addition for viral replication in the host.