A needle-free vaccine: what if we could vaccinate by massaging the skin?

The prospect of needle-free vaccination first and foremost addresses a well-documented barrier: the fear of injections.

In some individuals, this anxiety is sufficient to delay or even avoid a recommended vaccination. Yet this phenomenon is not trivial: it reduces vaccine uptake, creates access disparities and complicates large-scale campaigns, especially among children, highly sensitive individuals, and adults anxious about medical procedures. A needle-free vaccination technology could therefore overcome a major psychological barrier by turning an anxiety-inducing act into a simple, painless procedure.

The interest in a needle-free vaccine also fits within a context where aluminum salts, commonly used as adjuvants, continue to spark debate. Indeed, aluminum hydroxide and aluminum phosphate have been used for more than 90 years to optimize the immune response. They help retain antigens at the injection site and locally recruit immune cells, thereby prolonging antigenic stimulation. However, many fear that aluminum accumulates in the body and triggers allergies or proves toxic to the brain.

However, the efficacy and safety of aluminum in vaccines have been examined in dozens of publications, which concluded that it essentially causes transient local reactions : pain, redness, swelling, and sometimes a small nodule. No strong link has been established with lasting effects, autoimmune diseases, or an increased risk of allergies based on available data. The aluminum present in vaccines is eliminated through the same routes as that contained in food, where daily exposure is much higher: an adult ingests 7 to 9 mg per day, while a vaccine dose contains between 0.125 and 0.85 mg. Infants receive about 4.5 mg via their vaccination schedule, compared with 7 to 117 mg via their diet depending on whether they are breastfed, formula-fed, or fed soy-based milk. Despite this consensus, doubt persists among part of the population and fuels mistrust of vaccines.

Thus, the idea of a needle-free vaccine administered by massage could reduce two barriers: the fear and pain of injections, and the perceived risk associated with injected adjuvants like aluminum.

Recent research conducted by Élodie SEGURA’s team has highlighted the possibility of vaccinating by massaging the skin, demonstrating that stimuli mechanical stimuli influence the immune response. The researchers used a device capable of applying controlled stretch to the skin—equivalent to a therapeutic massage or the vigorous application of a cream. For twenty minutes, the skin of mice and human volunteers was subjected to this mechanical strain without causing visible damage. Observations showed that this stretching immediately altered the activity of epidermal cells, particularly keratinocytes, which respond to mechanical changes by releasing proinflammatory cytokines such as TNF-α, as well as neutrophils and monocytes, key immune cells. This response indicates that the skin interprets massage not as mere physical stimulation but as a potential danger signal capable of mobilizing local immunity.

However, it is the effect of massage on skin microarchitecture that constitutes the study’s most remarkable finding. The researchers observed that stretching induces a transient opening of hair follicles, as measured by both imaging and by the ability of fluorescent macromolecules to penetrate the skin (in this case dextran, a glucose polymer). Under normal conditions, follicles are relatively closed structures that limit the entry of bulky compounds. However, under the influence of massage, their opening increases enough to allow large molecules, far larger than those that typically cross the skin barrier (< 500 Da) to enter the follicular canal. Human skin, like that of the mouse, then becomes temporarily permeable to antigens applied topically. Notably, this increased permeability is transient, with the researchers observing that both mouse and human skin returned to their initial structure after several tens of minutes.

The study also revealed that this follicular opening not only permits the passage of exogenous antigens: it also facilitates the penetration of compounds derived from the skin microbiota, thereby activating dermal dendritic cells. This concept was supported by an increase in genes associated with pathogen-induced maturation, such as Cd86, Cxcl9, Cxcl10, and Myd88. In germ-free mice—i.e., those without a microbiota—stretching did indeed enhance neutrophil infiltration, indicating that this response does not depend on the microbiota. However, stretching neither recruited monocytes nor monocyte-derived macrophages, nor did it accelerate dendritic cell migration to the lymph nodes. In normal mice lacking hair follicles, stretching this time induced infiltration by neutrophils, monocytes, and macrophages, but without boosting dendritic cell migration.

These results show that stretching alone is sufficient to recruit neutrophils and monocytes, but that full activation of dendritic cells, including their migration to lymph nodes, requires the penetration, via hair follicles, of microbiota-derived molecules.

Vaccinating by massage or by injection engages different mechanisms. Intramuscular injection introduces the antigen into tissue that is sparsely populated with immune cells, which explains why injected vaccines often need to be combined with adjuvants, such as aluminum salts, to recruit and activate local dendritic cells. The immune response is then initiated as follows: the injection causes a micro-injury, the adjuvant creates an inflammatory site, and together they generate a sufficient signal to trigger dendritic cell maturation and their migration to the lymph nodes.

Vaccination by massage is different. The skin is naturally rich in dendritic cells and lymphocytes, which facilitates the immune response. When the antigen is deposited on the skin and then "pushed" into the follicles by massage, follicular dendritic cells can immediately capture and process the antigen. Unlike muscle tissue, the skin does not need to be "awakened" by an adjuvant, as mechanical stress and the entry of microbial fragments serve as immune activation signals.

Researchers then verified that the transient opening of hair follicles and the migration of dendritic cells induced by skin stretching could be exploited to administer a vaccine to mice. For this, they combined an H1N1 antigen (influenza vaccine) with a QS-21 adjuvant encapsulated in nanoliposomes, then used a fluorescent tracer to demonstrate that a single stretch of skin enabled efficient, noninvasive penetration of the vaccine into the epidermis and dermis.

The nanoliposomes penetrated the skin, then gradually released their contents into the bloodstream. The HA antigen also reached the draining lymph nodes, indicating active transport by dermal dendritic cells. The researchers then compared administration via skin stretching with an intramuscular injection containing the same antigen dose. They observed that the stretching method led to greater antigen accumulation in the lymph nodes and a higher anti-HA IgG response.

These results suggest that massage allows for effective, needle-free, non-invasive vaccination in mice.

Although this study provides intriguing evidence showing that skin stretching can enhance the penetration of macromolecules and activate certain immune cells through the transient opening of hair follicles, several questions remain unanswered.

One might first question the tolerability of this procedure in sensitive skin, which could respond poorly to 20 minutes of vigorous massage. Moreover, one may wonder whether a mechanical alternative such as microneedling—allowing controlled disruption of the skin barrier without 20 minutes of massage—would represent a simpler strategy to implement.

Moreover, the transcriptomic analyses performed have not precisely identified which cell populations respond to this stimulation: it is still unknown whether keratinocytes, dermal fibroblasts, or other stromal cells detect the stretch, nor which mechanoreceptors are involved. The exact role of the skin microbiota in this response also remains to be clarified, particularly the microbially derived molecules capable of triggering inflammation and activating immune cells. This question is all the more important given that the microbiota varies greatly from one individual to another, and that people with skin diseases (atopic dermatitis, psoriasis, rosacea…) often exhibit dysbiosis that could alter the immune response induced by massage.

Additionally, the researchers observed a rapid influx of innate immune cells within 24 hours after stretching, but neither their fate nor their contribution to dendritic cell activation was monitored over time. This lack of long-term immune follow-up therefore does not allow assessment of sustained protection or identification of potential late adverse effects.

The toxicological implications highlighted by the study also represent a major concern. The demonstration that massage transiently opens the follicles and increases the penetration of macromolecules means that this route could also facilitate the entry of undesirable substances, such as atmospheric pollutants or allergens. If this technique is performed incorrectly, it could trigger undesired immune responses. Furthermore, it remains to be determined whether this route of administration would induce systemic side effects similar to those sometimes observed after a conventional vaccination, such as fever or muscle aches.

Furthermore, one might question the types of vaccines compatible with this approach. In the study, the example used was an inactivated virus vaccine (H1N1), whose viral particles can penetrate into the skin via open hair follicles and be taken up by local immune cells. Conversely, it is uncertain whether this method is suitable for live attenuated vaccines, whose viral particles might be ineffective if the skin environment does not permit their replication. Similarly, mRNA vaccines, which are highly sensitive to stability conditions and require precise intracellular delivery, might not be compatible with topical application.

Moreover, the question of the absorbed antigen dose represents another key consideration. The study shows that in mice, massage can induce a qualitative immune response following application of an H1N1 vaccine. However, the exact amount of antigen actually captured by cutaneous dendritic cells has not been precisely quantified, and vaccine efficacy often relies on strict quantitative thresholds. Indeed, an insufficient dose could result in a weak or heterogeneous response. This limitation is all the more critical since the study does not assess the functional intensity of the response (antibody levels, viral neutralization, memory immunity, etc.).

Finally, if the study confirms that stretching increases the penetration of macromolecules into human skin, other experimental components conducted in mice cannot be reproduced as-is in humans. Given the significant differences between human and murine skin, further research will be necessary to determine whether the immune activation induced by stretching and its vaccination potential can actually be translated to humans.

• BOS J. D. & al. The 500 Dalton rule for the skin penetration of chemical compounds and drugs. Experimental Dermatology (2000).

• EXLEY C. & al. Insight into the cellular fate and toxicity of aluminium adjuvants used in clinically approved human vaccinations. Scientific Reports (2016).

• ROGERS M. A. M. & al. The fear of needles: A systematic review and meta-analysis. Leading Global Nursing Research (2018).

• ANDERSSON N. W. & al. Aluminum-adsorbed vaccines and chronic diseases in childhood: A nationwide cohort study. Annals of Internal Medicine (2025).

• SEGURA E. & al. Transient skin stretching stimulates immune surveillance and promotes vaccine delivery via hair follicles. Cell Reports (2025).