Everything the Licence Required

On 12 April 1955 Thomas Francis stood up at the University of Michigan and told a press conference that the Salk poliovirus vaccine was between 80 and 90 per cent effective against paralytic polio. He had the results of a trial of 1.8 million children: 420,000 given the vaccine, 200,000 given a placebo, 1.2 million given nothing and watched.

The United States government licensed the vaccine within a day of the announcement. Accounts differ on whether it was the same afternoon or the following morning, and one of them attributes the speed to political pressure. Over the next fortnight five companies distributed about five million doses.

Thirteen days after the first children were injected, reports began arriving of polio in vaccinated children.

Why nobody waited

The haste is easier to understand with the numbers in front of you. Within a decade of Karl Landsteiner identifying the poliovirus in 1908, an epidemic in New York killed 2,400 people, most of them children, and left thousands more permanently disabled. By 1951 the United States was recording 59,000 cases a year. Offit quotes the summary judgement of the period: polio was second only to the atomic bomb as the thing Americans feared most.

The March of Dimes, which took its impetus in part from Franklin Roosevelt's own paralysis, ran the largest public fundraising campaign that had then been attempted, and from 1951 it was giving Salk 200,000 dollars a year, which at the time was an enormous sum for one laboratory. A country in that state does not wait a fortnight for paperwork.

Cutter

Every one of those first cases had received vaccine made by Cutter Laboratories, a family firm in Berkeley, California. Vaccine made by Wyeth caused some cases too. Cutter's vaccine contained live poliovirus that the inactivation process had failed to kill.

Within a month the first mass vaccination programme against polio in the world had been abandoned.

The casualty figures are worth setting out carefully, because they do not agree. One published account gives 40,000 cases of polio, 200 children left with varying degrees of paralysis and 10 dead. Another gives at least 220,000 people infected with live virus, including 100,000 who caught it from a vaccinated child rather than from a needle, 70,000 who developed muscle weakness, 164 severely paralysed and 10 dead. Both of those are reviews of the same book by the same author. They cannot be reconciled from the published record, and the fact that two careful readers of one account came away with figures that far apart is itself part of the story of how this event is remembered.

The lower and upper bounds agree on one number. Ten children died.

What actually failed

The inactivation. Salk's method used formaldehyde to kill the virus while leaving enough of its structure intact to provoke immunity, and it worked: three larger companies made safe vaccine to exactly the same protocol.

Cutter's process did not. The seed virus was the Mahoney strain, which is highly virulent. The preparation contained cellular debris, and virus particles sheltering inside that debris were not fully exposed to the formaldehyde. The safety testing that should have caught surviving live virus was inadequate. Communication with other scientists and with the government was poor.

None of that is exotic. It is a small firm without deep experience attempting a manufacturing process at the edge of what was then understood, and getting it wrong in ways nobody was looking for.

The sentence that matters

Here is the finding that the popular version of this story leaves out entirely.

Cutter Laboratories was doing everything the licensing authority required of it.

The requirements were not adequate to the thing being made. The inspectors did not detect the deficiency because the inspection regime was not designed to. Paul Offit, who wrote the standard account and is about as far from an opponent of vaccination as it is possible to be, exonerates Salk and concludes that the federal government, through its vaccine regulatory agency, was in the best position to have avoided the tragedy.

That is why the regulatory consequence was the right one. The United States built a proper federal system for regulating vaccines, and vaccination resumed in the autumn of 1955. The safety record of American vaccines since has been described as unmatched by any other class of medical product.

The consequence nobody designed

The legal consequence went the other way.

Sixty lawsuits followed. The first went to a jury, which reached a verdict with unusual consequences: Cutter had not been negligent in producing the vaccine, but it had breached an implied warranty that its product was safe. The company was liable although it had done nothing wrong.

That is the birth of liability without fault, and it did not stay in the vaccine aisle. It shaped American product liability for the next fifty years.

For vaccines specifically the effect was severe enough that Congress had to intervene. The National Vaccine Injury Compensation Program was created in 1986 to shield manufacturers from claims that the science did not support, because vaccines had become one of the first classes of medical product that lawsuits came close to eliminating.

Companies left the business anyway. In 1957 twenty six companies made five vaccines in the United States. By 2004 four companies made twelve.

The last turn

There is one more thing, and Offit puts it more bluntly than a summariser should.

The Cutter incident convinced scientists and the public that Salk's killed vaccine was dangerous. The United States moved to Albert Sabin's live attenuated vaccine, which was given by mouth, was cheaper to administer, and spread protection to contacts. It could also revert to virulence passing through the gut. Through the 1980s and 1990s it was still paralysing six to eight children a year, until a modified version of the Salk vaccine was brought back.

The disaster caused by a badly inactivated vaccine led, in part, to the adoption of a vaccine that was more dangerous.

Sources

• Michael Fitzpatrick, "The Cutter Incident: How America's First Polio Vaccine Led to a Growing Vaccine Crisis", Journal of the Royal Society of Medicine (more than 200,000 children in five states receiving vaccine whose inactivation was defective; the abandonment of the programme within a month; the figures of 40,000 cases of polio, 200 children with varying degrees of paralysis and 10 deaths; Offit's exoneration of Salk and his conclusion that the federal vaccine regulatory agency was in the best position to have avoided the tragedy; three larger companies producing safe vaccine to the same protocol while Cutter's inexperience went undetected by inspectors; the court ruling that Cutter was liable although not negligent and the wave of litigation that followed; the 1986 National Vaccine Injury Compensation Program; and Offit's observation that the incident led in part to the adoption of a more dangerous vaccine, with the oral vaccine still paralysing six to eight children a year in the 1980s and 1990s before a modified Salk vaccine was reintroduced).
• Peter McIntyre, "The Cutter Incident: How America's First Polio Vaccine Led to the Growing Vaccine Crisis", British Medical Journal (the 1954 trial of 1.8 million children with 420,000 vaccinated, 200,000 given placebo and 1.2 million given nothing; licensure the day after the results thanks to political pressure; five companies distributing about five million doses in two weeks; the first cases appearing thirteen days after the first doses and all having received Cutter vaccine, with Wyeth's also causing some; the figures of at least 220,000 infected including 100,000 contacts, 70,000 with muscle weakness, 164 severely paralysed and 10 dead; the Mahoney strain, the deficiencies in inactivation, the inadequate safety tests and the poor communication; the statement that Cutter was doing all that the licensing authority required; the sixty lawsuits and the verdict of no negligence but breach of implied warranty that created liability without fault; and the fall from 26 companies making five vaccines in 1957 to four companies making twelve by 2004).
• Centers for Disease Control and Prevention, "We Were There: Conquering Polio in America, the Cutter Incident and Beyond" (batches that passed the required safety testing while containing live virus; over 250 cases of polio attributed to one company's vaccine; the recall as soon as cases were detected; and the resumption of vaccination in the autumn of 1955 after federal oversight was improved).
• The College of Physicians of Philadelphia, "History of Vaccines: polio timeline" (Thomas Francis announcing at the University of Michigan that the vaccine was 80 to 90 per cent effective against paralytic polio, and the government licensing it later the same day).