Dementia and the Tuberculosis Vaccine: Reduced Tau Burden With Administration in Mice

Retrospective studies of patients with bladder cancer have associated treatment with Bacillus Calmette-Guérin (BCG) with a lower incidence of dementia, but could these findings be clinically applicable?¹ In a recent study, William R. Bishai, MD, PhD, and colleagues vaccinated PS19 tauopathy mice with BCG at 3 months of age, before symptom onset, and observed reduced tau neurofibrillary tangles.² Bishai noted that earlier BCG studies in amyloid-beta mouse models improved memory performance without changing lesion burden, whereas this work reduced tau neurofibrillary tangle burden. He concluded that BCG’s safety record left no real obstacles to prospective human trials and that, unlike other vaccines associated with dementia prevention, BCG is not used in the United States or western Europe, so it could offer protection beyond that of existing vaccines if proven effective.

Psychiatric Times: What led your team to study BCG in the context of dementia?

William Bishai, MD, PhD: My research career has focused on tuberculosis and improving the tools to combat that disease. This prompted a long-standing interest in the BCG vaccine, and trying to improve it by making recombinant BCG strains that would perform better at tuberculosis prevention as well as bladder cancer immunotherapy. Along the way, I became interested in the numerous off-target uses for BCG that have been identified in humans. As you may know, BCG has been used as an immunotherapy for many cancers well before its antitumor activity against bladder cancer was defined in 1975. Moreover, it also has been used against multiple sclerosis, type 1 diabetes, and as a protective vaccine for numerous upper respiratory infections.

Five years ago, I was fortunate enough to be invited to a symposium commemorating the 100th anniversary of the development of BCG held at the Institut Pasteur in Lille, France. This institute was where Calmette and Guerin performed the original development work to derive BCG that was first tested in humans in 1921. At that meeting, there was a session on “BCG off-target effects” with a talk that highlighted the potential role of BCG against dementia. I had a chance to speak more with the presenter of that fascinating talk. He told me that his mentor—a gentleman in his 80s who was sharp as a pin—has been growing BCG in the lab every few years and drinking 1×108 bacilli, and that he does so for its dementia protective effects. So when I got back from the meeting, I looked into the details, and lo and behold indeed there are numerous retrospective studies in bladder cancer cohorts that show statistically significant benefits in BCG recipients versus nonrecipients. And there are also some small prospective clinical trials of BCG that showed a beneficial effect with dementia biomarkers. These factors prompted us to begin the work described in our recent paper.

PT: How might the immune system contribute to the development of Alzheimer disease and related dementias?

Bishai: There are numerous hypotheses as to the underlying molecular basis of Alzheimer disease and related dementias. Aberrant biochemical processing of amyloid and/or tau proteins, lipid dysregulation including ApoE lipid transport, aberrant endosomal-lysosomal maturation, mitochondrial dysfunction, vascular disease, and even infectious etiologies have all been studied as potential causes.

Not surprisingly, dysfunctional immune activation has also been proposed as an etiologic mechanism. This might lead to impaired protein clearance with concomitant synaptic destruction, subsequent neuronal injury, and escalating inflammatory stimulation. In particular, microglial cells, known to be the brain’s resident immune cells, are closely implicated in clearing amyloid-beta and tau aggregates.Microglial hyperactivation, also called microgliosis, is a known hallmark of AD as well as a source of persistent neuroinflammation.

Since BCG is a potent stimulator of the immune system, our recently published work bolsters the concept that immune modulation may prevent Alzheimer-associated neurodegeneration. Importantly, with BCG showing a robust phenotype in the PS19 tauopathy mouse model, we are now equipped with a valuable tool to further investigate the specific mechanisms by which BCG function in this mouse model. In addition to the clues that these future studies may offer towards improving BCG as a preventive agent, the same studies may also be enlightening as to underlying mechanisms of neurodegeneration in AD and related dementias.

PT: Your mice had tau pathology and received BCG at 3 months of age. Why did you choose that timing, and what does it imply about when a vaccine might need to be given?

Bishai: We wanted to dose adult mice at the earliest possible opportunity, and 3 months of age mark is generally considered to be the onset of young adulthood in mice. Also, the PS19 tauopathy mice start showing symptoms at the age of 6 to 7 months, so we wanted to be sure to vaccinate them before the onset of symptoms.

We are currently doing ongoing experiments to see if BCG may have a therapeutic effect when given after the onset of symptoms to reveal if the vaccine may have a therapeutic effect above and beyond the preventive effect observed in the recent paper.

Regarding the use of BCG and humans, my personal opinion is that the mechanism of dementia prevention by BCG is related to trained immunity with epigenetic and metabolomic reprogramming of myeloid and microglial cells. Trained immunity in humans has been shown to wane after about 12 months. While this is speculative and not proven, it may be that if BCG does have a future as a human dementia vaccine it will need to be given annually like seasonal flu vaccinations.

PT: What do you think is driving the reduction in tau deposits in treated mice?

Bishai: Our study showed that microglial cells in vaccinated, tauopathy-prone (PS19) mice showed a higher degree of phagocytic activity, and that they removed tau peptides from the extracellular space more rapidly than those from unvaccinated mice. Hence, we think that BCG-mediated stimulation microglial cells is a pivotal part of the mechanism by which the vaccine reduces tau neurofibrillary tangles.

PT: How did you determine that "trained immunity" is a likely mechanism rather than a general anti-inflammatory effect?

Bishai: In the present study we focused on whether BCG had efficacy for preventing tauopathy progression in the PS19 mice. But we still have a ways to go to nail down the specific mechanism.

We hypothesize that trained immunity is at play because there were a few clues in our study. First, we found that microglial cells from the brains of BCG vaccinated tauopathy-prone (PS19) mice had a metabolic shift toward glycolysis, which is one of the hallmarks of trained immunity. And a second clue was that we also observed a higher level of peripheral monocytes expressing a known trained immunity surface marker, CD80.

Our current studies are intended to better understand the precise cellular mechanisms responsible for the effect. One important characteristic of trained immunity we will look for is epigenetic reprogramming of myeloid and microglial cells with DNA demonstrating classic activation marks. These include the H3K4me3 and H3K27ac epigenetic marks.

PT: What role do myeloid cells in the brain appear to play in clearing tau tangles?

Bishai: CNS microglial cells are of myeloid origin. Our work showed that while both vaccinated and unvaccinated tauopathy-prone (PS19) mice have elevated levels of microglial cells in their brains compared to healthy control mice, BCG vaccination led to reduced levels of deleterious activation— or microgliosis—among these microglial cells. We were also able to show that isolated microglial cells from vaccinated mice showed higher levels of phagocytic activity and could specifically and engulf and remove tau peptides.

A second category of myeloid cells that we observed changes in were in CNS monocytes recruited from the periphery. Our work showed that brain monocyte recruitment occurred at a significantly higher levels in BCG vaccinated mice than in the unvaccinated animals. Although we have yet to prove this particular mechanism, it is possible that BCG-mediated recruitment of peripheral monocytes into the CNS significantly augments the ability of endogenous CNS microglial cells to remove tau neurofibrillary tangles.

PT: How do your tau model results compare with earlier BCG work in amyloid models?

Bishai: One of the reasons we focused our work on the mouse tauopathy model with PS19 mice was that, as you point out, there are prior studies of BCG in preventing amyloid-beta disease in a different line of transgenic mice including at least 3 studies from Chinese laboratories. To my knowledge, we are the first to look at BCG in dementia prevention using a tau mouse model.

Similar to our work, the amyloid-beta mouse models showed improvement in mouse performance in memory behavioral tests following BCG vaccination.

An important difference between the 2 models is that our work shows that BCG reduces the tau neurofibrillary tangled burden, whereas the amyloid-beta studies showed no change in CNS lesion burden with BCG. Another difference was that the amyloid-beta studies showed that BCG led to a higher level of anti-inflammatory cytokines. In contrast, our work identified brain reprogramming of microglial cells as showing the most significant change mediated by BCG vaccination.

PT: What questions need answering before BCG is tested in people at risk for dementia?

Bishai: BCG was first used in humans in 1921. It is thought to be the most widely used vaccine in human history.At least 340 million infants receive BCG annually and several hundred thousand adults are treated with BCG annually for bladder cancer. BCG is also well tolerated, and adverse events are generally mild and occur infrequently.

Not surprisingly, with this excellent safety record, BCG is often used in human studies assessing fundamental aspects of human immunology including trained immunity. There are also numerous human studies of BCG assessing its role in off-target diseases, such as type I diabetes, multiple sclerosis, and prevention of viral infections,

So, there are no real obstacles for prospective clinical trials in humans to assess the efficacy of BCG in dementia prevention. Indeed, one such study is ongoing in Israel. Moreover, there is a US-based consortium—which I pleased to be a member of—that has proposed a large prospective clinical trial to test the efficacy of BCG for the prevention of dementia in older, at-risk adults.

PT: What 1-2 items would you highlight for the practicing psychiatric clinician?

Bishai: My first point would be that our study adds to the growing evidence that commonly used vaccines may prevent dementia. My second point is that most of the vaccines associated with dementia prevention are already recommended for broad use in the United States population. However, BCG is unique in that United States, western Europe and other wealthy countries do not use it. Hence, should BCG ultimately be proven effective against dementia, it would offer unique prevention activity above and beyond what most of the population already gets from the other vaccines.

References

1. Weinberg MS, Zafar A, Magdamo C, et al. Association of BCG vaccine treatment with death and dementia in patients with non–muscle-invasive bladder cancer. JAMA Netw Open. 2023;6(5):e2314336.

2. Study advances mechanistic insight into tuberculosis vaccine’s link to reduced dementia risk. Press release. October 21, 2026. Accessed October 26, 2026. https://www.hopkinsmedicine.org/news/newsroom/news-releases/2026/10/study-advances-mechanistic-insight-into-link-between-tuberculosis-vaccine-and-dementia-risk-reduction