Questioning Offit. Part 9. Pneumococcus and rotavirus

Previous parts:
Part 1: Introduction
Part 2: DTP
Part 3: Encephalopathy
Part 4: Dravet syndrome
Part 5. SIDS
Part 6. West syndrome
Part 7. Hib
Part 8. Hepatitis B

One of the saddest lessons of history is this:
if we’ve been bamboozled long enough,
we tend to reject any evidence of the bamboozle.
Carl Sagan

Chapter 5 (continued)

Pneumococcus

After the vaccines against Hib and hepatitis B, Offit writes, Barbara Loe Fisher chose her next target: the pneumococcal vaccine.

In 1998, researchers in Northern California performed a landmark study. They tested the pneumococcal vaccine in thirty-eight thousand infants: half received it and half didn’t. The results were dramatic. Pneumococcus caused bloodstream infections in seventeen children who didn’t get the vaccine and in not a single child who did: seventeen versus zero. Excited, researchers made their findings public.

Sounds effective. The question is whether it's just as safe. This is the first time in the book that Offit cites a randomized clinical trial (RCT), so let’s take a closer look at it.

In the vaccinated group, 4 infants died from sudden infant death syndrome (SIDS), while in the control group, there were 8 such deaths.
In the vaccinated group, 1,188 infants visited the emergency department within 30 days, compared to 1,169 in the control group.
In the vaccinated group, 513 infants were hospitalized within 60 days after vaccination, compared to 579 in the control group.

Sounds pretty safe. It seems that the pneumococcal vaccine causes little to no serious side effects, since the differences between the vaccinated and control groups are minimal. Moreover, it even appears that the pneumococcal vaccine may reduce the risk of sudden infant death syndrome.

But that’s only until we find out that the placebo given to the control group was in fact a meningococcal type C vaccine. And the meningococcal type C vaccine had never been licensed in the U.S., where this study was conducted — meaning that the control group received an experimental meningococcal vaccine.

The authors report that one infant died of SIDS within a week of receiving the pneumococcal vaccine, and two died after receiving the meningococcal vaccine. They write: “In an age- and seasonality-adjusted analysis based on the California State SIDS data, we would have expected 1.06 cases within 1 week of either vaccine.”

However, the authors fail to emphasize that the study included only healthy infants. Children with immunodeficiencies or any serious medical conditions in their history — in other words, those at higher risk — were excluded from the outset. Accordingly, for a sample consisting entirely of healthy infants, the expected mortality rate should have been significantly lower than the California average.

More than 6% of completely healthy infants visited the emergency department within a month after vaccination (that’s one in every 16 participants). About 3% of them were hospitalized within two months (roughly one in every 34 participants). Is that a lot or a little? Let’s do the math.

According to a 2012 study, about 10% of infants in California are hospitalized during their first year of life. That averages out to roughly 0.8% per month (10% divided by 12). That’s an upper limit — the real number is likely lower, since most hospitalizations occur in the neonatal period. So the upper bound for hospitalization over a two-month period among all infants — both healthy and ill — is about 1.6%.

Now compare that to the pneumococcal vaccine trial: among carefully selected, perfectly healthy infants, the hospitalization rate over the same period was 2.9%. And the researchers weren’t bothered by this at all. What’s more, the rate of hospitalizations (unrelated to pneumococcus) among those who received the pneumococcal vaccine was statistically significantly lower than in the group that received the meningococcal vaccine. That didn’t raise any eyebrows either. Based on this data, the vaccine was declared completely safe.

Offit:

As she had done with the Hib vaccine, Fisher allied herself with Bart Classen, who, without any supportive evidence, claimed that the pneumococcal vaccine caused diabetes. Also, Fisher had claimed that diseases caused by pneumococcus were not particularly common, affecting only those in poor health. This wasn’t true. Before the vaccine, every year pneumococcus caused four million ear infections, one hundred and twenty thousand cases of pneumonia requiring hospitalization, thirty thousand bloodstream infections, and twenty-five hundred cases of meningitis. Most of these diseases occurred in previously healthy children. And, although the bacterium didn’t kill as many children as measles or polio, John McKenzie’s comment that it killed “relatively few, only about two hundred” was rather callous to those parents whose children had suffered and died from pneumococcus.

By 2009, almost a hundred million doses of the pneumococcal vaccine had been given to American children. As a consequence, the incidence of pneumococcal disease has decreased dramatically. Far fewer children now get meningitis, pneumonia, and bloodstream infections caused by pneumococcus. And, although we’ll never know their names, hundreds of children are still alive because they got the pneumococcal vaccine.

First of all, if a disease so rarely results in death—200 cases per year in the U.S., or roughly 1 in 20,000 children—it might be worth evaluating the potential impact of the vaccine on mortality a bit more carefully, to be sure the vaccine isn’t more dangerous than the disease itself. In the study mentioned above, four completely healthy children died from SIDS alone (the authors do not report how many died from other causes). That means roughly 1 in 4,700 infants vaccinated against pneumococcus died suddenly for unknown reasons. Whether these deaths were related to the vaccine, we will never know—because the study included no control group of completely unvaccinated children. Without studies using an inert placebo, no one can reliably claim that the vaccine saves more lives than it may cost.

Second, I’m not disputing that the incidence of invasive pneumococcal infections declined following the introduction of vaccination. However, surveillance for these infections in the U.S. only began several years after mass vaccination against Hib had already started. According to a Finnish study, between 1992 and 1994, the incidence of pneumococcal infections doubled among children under two years old and tripled among children up to age 16. The authors linked this to the start of Haemophilus influenzae vaccination. A similar increase in disease incidence was observed in Philadelphia.

Haemophilus influenzae, pneumococcus, and meningococcus are part of the normal flora of the nasopharynx and typically do not cause illness on their own. These are so-called opportunistic infections—they become dangerous only in the context of a weakened immune system. Therefore, eliminating one type of bacteria through vaccination leads to its replacement by other strains or entirely different bacteria. For example, the displacement of pneumococcus from the nasopharynx leads to its replacement by Staphylococcus aureus. [1] This happens because pneumococcal bacteria produce hydrogen peroxide, which kills staphylococcus bacteria. [1] Similarly, pneumococcus suppresses the growth of meningococcal bacteria.

There is a complex balance among the bacterial species inhabiting the nasopharynx. Eliminating one species—even if it reduces the number of cases it causes—does not necessarily result in a drop in overall illness. For instance, a 2006 study found that while pneumococcal meningitis incidence dropped threefold after vaccination began, the total number of bacterial meningitis cases remained unchanged, and cases of bacteremia actually increased.

Offit does not address any of this. After citing just two studies—neither of which proves the pneumococcal vaccine’s safety—he moves on to the next topic:

Because the vaccine has reduced the number of children who carry pneumococcus in the nose and throat, older people, such as their grandparents, have also benefited. Finally, the chronic diseases predicted by Barbara Loe Fisher and Bart Classen in front of millions and millions of television viewers never materialized.

According to a study published in 2011, more than half (!) of children in the U.S. have at least one chronic condition—and that number continues to rise. It’s strange that Offit doesn’t seem to notice this.

As for the claim that vaccination reduced the number of children carrying pneumococcus—that’s true, but it does not necessarily follow that this benefited their grandparents. Vaccination leads to the replacement of one pneumococcal strain with another, and no meaningful reduction in the incidence of invasive infections among the elderly has been observed. That’s precisely why, since 2014, the CDC has recommended pneumococcal vaccination for older adults. However, this recommendation was not based on data showing the vaccine’s effectiveness in preventing illness, but solely on studies of immunogenicity—that is, the vaccine’s ability to generate antibodies. But does the presence of antibodies after vaccination mean the vaccine is effective?

In a large prospective study published in 2020, which followed more than 2 million people over age 50 in Spain for two years, the conjugate pneumococcal vaccine (Prevnar-13) was found to be associated with a 52% increase in the risk of pneumococcal pneumonia, and a 72% increase in the risk of pneumonia from all causes. Another pneumococcal vaccine (Pneumo 23) was associated with a 17% increase in the risk of pneumonia. The authors concluded that pneumococcal vaccines are ineffective in adults. The vaccines were also found to be ineffective in high-risk groups for whom they are officially recommended. The authors mention a randomized trial published in 2015, which showed a 46% effectiveness rate for Prevnar-13 in the elderly. However, the placebo used in that study contained aluminum and polysorbate 80. The increased risk of pneumonia following Pneumo 23 vaccination has also been confirmed in four other studies—two randomized trials and two cohort studies [1] [2] [3] [4].


Rotavirus

In 1998, Offit writes, an event occurred that should have, at least in theory, dispelled any lingering doubts Barbara Loe Fisher had about the government’s genuine concern for vaccine safety.

In August 1998, the first vaccine against rotavirus was licensed—a disease that caused approximately 60 child deaths annually in the U.S., mostly due to dehydration.

Offit:

Because the disease is common and occasionally fatal, the CDC recommended the vaccine for all infants...
In July 1999—ten months after licensure—the CDC discovered something it hadn’t anticipated. Reports of fifteen children who had received rotavirus vaccine appeared in the Vaccine Adverse Events Reporting System (VAERS). These children shared several features: all had developed an uncommon form of intestinal blockage called intussusception (which occurs when one segment of the small intestine telescopes into another and gets stuck), all had recently received the rotavirus vaccine, and most were about two months old (intussusception is unusual in two-month-old children). Intussusception is a medical emergency. Children with the disorder can develop severe bleeding from the intestine or a bloodstream infection; both can be fatal.
The CDC knew that it had a problem. So, on July 16, 1999, it temporarily suspended the use of the new rotavirus vaccine until investigators could figure out what was going on. Finding an answer wasn’t going to be easy. Every year before the rotavirus vaccine had become available, one in two thousand infants in the United States developed intussusception, most between five and nine months of age. Jeff Koplan, head of the CDC, pulled people off other projects, spent millions of dollars, and made it clear that an explanation for the intussusception found in two-month-olds must come quickly. By October, only months after the possible association between rotavirus vaccine and intussusception was reported to VAERS, Koplan and his CDC team figured it out. Children who received the new rotavirus vaccine were twenty-five times more likely to get intussusception than those who hadn’t. Although the rotavirus vaccine clearly caused intussusception, the risk of getting the disease was quite low: one case per ten thousand vaccine recipients. That same month, the CDC withdrew its recommendation of the vaccine—and the company that produced it took it off the market. Seven years would pass before a safer rotavirus vaccine was made. The serious side effect caused by the first rotavirus vaccine offered several insights into how the CDC monitors vaccines... This is what happens when a vaccine actually causes a problem. Barbara Loe Fisher should have been reassured by all of this. But she wasn’t.

Tears of gratitude toward the CDC involuntarily well up when reading this passage. What a blessing it is to have an organization that cares so deeply about children's health!

But let’s take a closer look at how truthful this story really is. Were there any signs, prior to the vaccine’s approval, that might have indicated a possible link between vaccination and intestinal intussusception? Was this complication truly, as Offit claims, a complete surprise?

In the clinical trials of the RotaShield vaccine, intussusception developed in 5 out of 10,000 vaccinated infants (i.e., 1 in 2,000), compared to 1 in 4,600 in the control group. Although this difference did not reach statistical significance, something else was far more important: in three of the five vaccinated infants, the intussusception occurred within the first week after vaccination. That result was statistically significant, and as a consequence, the manufacturer listed intussusception as a potential side effect directly in the vaccine’s package insert.

In other words, the CDC’s working group spent millions of dollars and several months of investigation to confirm what had already been written in the vaccine insert from the very beginning. The dried-up tears of gratitude well up again—but this time, in gratitude to Barbara Loe Fisher, who, unlike the CDC, actually read the package insert.

That’s about all Offit tells us about the rotavirus vaccine. He modestly omits the fact that he himself is one of the developers of the RotaTeq vaccine, which was licensed seven years after the events he describes—and that he earned millions of dollars from the sale of the patent for that vaccine. Additionally, Merck, which manufactures this vaccine, pays him over a million dollars a year for a position at a children's hospital that was created specifically for him.

Offit does not mention that in 1998 he was a member of the ACIP—the Advisory Committee on Immunization Practices—that voted to include the RotaShield vaccine in the childhood immunization schedule. Offit personally voted in favor. At the time, he was also holding a patent on his own rotavirus vaccine, yet he did not consider this to be a conflict of interest. However, a year later, when the committee reviewed and ultimately withdrew its recommendation, Offit abstained from voting—this time deciding that a conflict of interest did, in fact, exist.

He also fails to mention that, according to official statistics, there had not been a single recorded case of a child dying from rotavirus in the U.S. prior to the development of rotavirus vaccines in the 1990s. Only once work on the vaccines had begun did the CDC start asking the question: how many children in the U.S. actually die from rotavirus infection? Officially, not one child had died from it. Since no fatal cases had been registered, CDC researchers conducted their own studies and concluded that rotavirus might be responsible for somewhere between 20 and 60 child deaths annually. However, how accurate those estimates are remains unknown. No subsequent study has ever demonstrated that the rotavirus vaccine reduces childhood mortality.

Offit also doesn’t disclose that the two newer rotavirus vaccines—including RotaTeq, which he helped develop—can also cause intestinal intussusception, although at a significantly lower rate than RotaShield. The RotaTeq vaccine has been associated with a ninefold increase in the risk of intussusception—meaning that 1 in every 65,000 vaccinated infants may develop the condition. Is that a lot or a little? Considering the CDC’s own estimate that only 1 in 200,000 infants dies from rotavirus, the risk of a potentially fatal complication from the vaccine is three times higher than the risk of death from the disease itself. That doesn’t sound quite so negligible.

In 2010, independent researchers accidentally discovered the presence of porcine circovirus in the Rotarix vaccine, prompting the FDA to suspend its use. Initially, the FDA stated that RotaTeq did not contain the swine virus, but two months later, it was revealed that RotaTeq did, in fact, contain DNA from two types of porcine circoviruses. An FDA-convened committee concluded that these viruses were most likely harmless to humans and that the potential benefits of vaccination outweighed the possible risks. The committee also recommended that manufacturers develop vaccines free of swine viruses. One week after the discovery of the viruses in RotaTeq, the FDA recommended continuing the use of both vaccines. More than 15 years have passed since then, yet manufacturers have shown little urgency in developing virus-free versions.

Since rotavirus diarrhea—though unpleasant—is rarely life-threatening, the key question is whether vaccination reduces mortality. A Cochrane systematic review analyzing 60 studies found no evidence that rotavirus vaccination reduces mortality—neither in high-income countries nor in low-income ones.

In a randomized clinical trial of the Rotarix vaccine, significantly more vaccinated infants died of pneumonia compared to those in the control group. The control group received the same vaccine, but without the viral component. In the documents submitted by the manufacturer to the FDA, this increased risk was not deemed statistically significant. However, a recalculation by FDA experts themselves showed that the risk was statistically significant (see p. 129). Nevertheless, this did not prevent the FDA from licensing the vaccine.

More detailed posts on pneumococcus and rotavirus.

In the next part, we'll talk about the HPV vaccine, which will conclude our review of Chapter Five.