by Gertrud U. Rey
Vaccination with the AstraZeneca and Johnson & Johnson adenovirus‑based COVID-19 vaccines has been linked to a very rare but serious adverse event known as vaccine‑induced immune thrombotic thrombocytopenia (VITT). Based on this small but significant risk, the U.S. Food and Drug Administration revoked the emergency use authorization for the Johnson & Johnson vaccine in 2023, while the AstraZeneca vaccine was never authorized in the United States. Until recently, the cause of VITT was poorly understood, but a new study has clarified the biological mechanism.
VITT combines two conditions that appear contradictory. Thrombocytopenia is characterized by an abnormally low count of platelets – small blood components that normally stop bleeding by forming clots – so reduced platelet levels therefore cause bleeding risks due to poor clotting. In contrast, thrombosis involves excessive clot formation within blood vessels, which can obstruct blood flow and damage organs. The simultaneous occurrence of both thrombocytopenia and thrombosis in VITT is caused by immune-mediated activation of platelets, which drives extensive clot formation and depletion of circulating platelets faster than they can be replenished. This condition results in dangerous clotting and critically low levels of platelets.
Previous research had shown that people with VITT generate antibodies against platelet factor 4 (PF4), a protein that promotes clotting at sites of injury. These antibodies erroneously bind PF4 to form large immune complexes that trigger widespread platelet activation, leading to clot formation in unusual locations, while simultaneously depleting platelets. However, key questions remained unanswered: why do these antibodies appear in the first place, and why only in exceedingly rare individuals?
The authors of the new study observed that adenoviral core protein pVII shares important biochemical features with PF4, suggesting that antibodies raised against pVII during adenoviral vaccination – or even during natural adenovirus infection – could potentially cross‑react with PF4 in an autoimmune mechanism known as “molecular mimicry.” To test this hypothesis, the researchers isolated anti‑PF4 antibodies from people who developed VITT after vaccination, as well as from individuals with natural adenovirus infections. They then analyzed these antibodies in detail, examining their amino acid sequences, structures, and binding behaviors.
The results were striking. Adenovirus-specific antibodies from VITT patients bound both pVII and PF4, whereas anti-pVII antibodies from healthy vaccine recipients – who never developed VITT – did not bind PF4.
The study further showed that molecular mimicry alone was insufficient to trigger VITT; instead, two highly specific conditions were required. First, all individuals who developed VITT generated PF4-binding antibodies that were derived from one of two specific antibody gene sequences common to all affected individuals. Second – and more critically – these gene sequences acquired a very specific mutation that resulted in a substitution of a negatively charged amino acid at a precise location in the antibody light chain. Each antibody that had this amino acid change could simultaneously bind two PF4 molecules, leading to PF4 clustering and the formation of large immune complexes capable of activating platelets.
The authors confirmed the relevance of this amino acid change by reversing the mutation. When the negatively charged amino acid of the antibody was changed back to the original amino acid, the antibody no longer bound PF4, but instead preferentially bound pVII.
VITT is yet another example of virus-induced autoimmunity, in which the immune system mounts a strong response against a viral protein, only to accidentally cross a boundary and target a protein from the host. Similar mechanisms have been identified in autoimmune diseases linked to Epstein-Barr virus, such as multiple sclerosis and lupus (discussed previously here and here).
Rather than undermining confidence in vaccination, this study illustrates the power of biomedical science to identify problems and use that knowledge to engineer safer solutions. Armed with these insights, adenovirus-based vaccine vectors could be readily improved by replacing pVII with viral proteins that do not stimulate problematic B‑cell responses. One might reasonably ask whether this vaccine platform remains necessary given the availability of alternative technologies. However, adenoviral vaccines continue to play an important global role: they are relatively inexpensive, quick to manufacture, and easier to distribute than mRNA vaccines – making them particularly valuable in low‑ and middle‑income countries.
[This paper was also discuss on TWiV 1299.]