Skip to Main Content

It all started with a bit of bacteria swiped from the udder of a tuberculous French cow. Initially, it was just another livestock sample, a speck of virulence lounging in a Paris lab. But then it began to change. Two scientists had brought some of it to Lille, near the Belgian border, and had fed it a concoction of potato, glycerin, and ox bile. The bugs liked this witches’ brew, and multiplied, and multiplied again.

After years away from live animal hosts, the bacterium got used to its cushy lab-dish life and lost its barnyard edge. Enough of its tough old self remained to trigger a body’s immune defenses, but not enough to make a healthy person sick. Perfect for a microbial training drill. Word spread. Scientists made pilgrimages to France, to take home some defused bovine tuberculosis of their own — and what began as an udder-swab in 1902 ended up yielding one of the world’s most widely used vaccines, now given to some 100 million infants globally each year.


Because of this history, what we consider a single vaccine is in fact many, made thousands of miles apart. They’re like siblings, descended from the same stock, sharing all sorts of quirks and genes. But they aren’t quite identical. Now, a study published Tuesday has carefully catalogued how many live, but weakened, bacteria the vials from different vaccine manufacturers contain — and some turn out to have 10- to 1,000-fold more than others when grown up in the lab. You might think that that kind of whopping difference would mean some makers’ products are more protective than others, but researchers simply aren’t sure.

“In what world is it acceptable to have such variance in a drug or vaccine that we’re giving to infants?” said Dr. Ofer Levy, director of the precision vaccines program at Boston Children’s Hospital and a lead author on the study. “Imagine we marketed a heart medication and some people were getting 10- to 1,000-fold more of the active ingredient than others,” he added. “It would be an outrage!”

Yet these fluctuations were only observed in the lab, and so it’s hard to know their real-world implications. “Whether that means anything clinically — there isn’t evidence one way or the other,” said Dr. Punam Mangtani, a professor of clinical epidemiology at the London School of Hygiene and Tropical Medicine, who wasn’t involved in the research.


Filling in the gaps would be hard, though. Tuberculosis is the earth’s number one infectious killer, but primarily an illness of the poor. Its researchers and policymakers are in the unenviable position of dealing with problems that are at once urgent and underfunded. Whether these variations should be a priority is a matter of debate.

The vaccine is called bacille Calmette-Guérin — or BCG — after the rod-shaped bacteria it contains and the two microbiologists who first developed it. Since it was first tested in humans in 1921, it’s been given over 4 billion times. Studies on its effectiveness have shown conflicting results, but even so, epidemiologists estimate that every year those weakened microbes save thousands upon thousands of lives.

That BCG is not a single standardized vaccine is baked into the story of how the global supply came to be. When French doctors first saw, in the 1920s, that babies who ingested their concoction were less likely to die of TB, international scientists began showing up, asking for a smidge of the secret sauce. They didn’t all jump at once. The Russians came in 1924, the Danes in 1931. What each team took home wasn’t exactly the same, though. As Dr. Marcel Behr, who was the founding director of the McGill International TB Center, explained, “the thing that was being given out was a moving target, because the bacteria was continually propagated. It was changing and mutating in the lab.”

That didn’t stop once the vaccine arrived at its various destinations. Even if everyone fed the bugs potatoes, the potatoes in Moscow weren’t identical to those in Montreal. So, like an heirloom apple, the taste and texture dependent on soil and water and growing technique, the bacteria shifted slightly according to each producer’s recipe. “It’s an artisanal vaccine,” said Behr.

Researchers have known that for a long time. They’ve also known that the international standards for variance between different kinds of BCG are looser than they would be for a newer vaccine. The different formulations are just as safe as each other if you have a healthy immune system, but they might not be made exactly the same way.

What prompted Levy and his colleague, neonatologist Dr. Asimenia Angelidou, to compare different manufacturers’ versions of BCG was a blip of studies suggesting that the vaccine doesn’t just have positive effects in protecting against TB; it also might give a more general boost to the body’s defenses. But those results weren’t stable from place to place.

“The liveness, the amount of live mycobacteria per dose of the vaccine, might be an important parameter in how the vaccine acts,” explained Levy. “So I said, ‘Hey, why don’t we directly compare how many live organisms there are?’”

They ordered BCG from Denmark, Bulgaria, Japan, India, and the U.S., reconstituted the freeze-dried bacteria, and began growing them in the lab. Not only did they find that the Indian and Bulgarian versions — both of which are descendants of the stuff scientists originally took back to Russia — contained many fewer bacteria than their Japanese and American counterparts, they also found them more finicky, easier to cultivate with some chemicals than others. That’s why the team reported that these vaccines grew between 10 and 1,000 times less than others: The numbers of bacteria changed depending on the environment in which the Indian and Bulgarian vaccines were grown.

When the team put the vaccines in with umbilical cord blood donated after cesarean sections, they saw differences in the immune cells that were generated as a response — though it’s not entirely clear what sort of reaction is needed for the vaccine to protect against certain kinds of TB.

To Levy, now’s the time to pick apart the effects of these different vaccines in humans. “This is a call to action,” he said, “to scrutinize much more carefully the production and the standardization of this vaccine.”

Easier said than done. BCG is old and imperfect, better at preventing tuberculosis infections in organs other than the lungs, and more effective in babies than in adults. Figuring out how different manufacturers’ versions differ in their ability to prevent the disease would take a gargantuan effort that some feel would do more good if geared toward continuing the development of newer, potentially better vaccines. In the last few years, one candidate, made by GSK, was tried in 3,300 adults who already had TB lying dormant in their bodies. Among those who got the experimental vaccine, the number who became sick and contagious was half what it was among the unvaccinated — hardly ideal, but of enormous potential benefit worldwide.

Still, BCG is the only TB vaccine currently on the market. Many feel that truly understanding the variations between different manufacturers’ products is important but unlikely. “We would have to do an enormous clinical trial where we compare these vaccines head to head,” said Dr. Thomas Scriba, a professor at the University of Cape Town and deputy director of immunology at the South African TB Vaccine Initiative. “It would involve thousands of people, and cost an enormous amount of money, no one’s willing to do that study. Well, the scientists are willing to do the study. I cannot think of a single funder that would say yes, we would want to put $150 million into this trial.”

It’s a sentiment echoed by Dr. Johan Vekemans, a medical officer at the World Health Organization who works on tuberculosis vaccines. “You have to understand that BCG is a relatively cheap vaccine,” he said. “There’s going to be little appetite for providing more investments into the science of neonatal BCG — even if the WHO calls for more investigation.”