Immunologic mechanisms of seasonal influenza vaccination administered by microneedle patch from a randomized phase I trial

Seasonal influenza causes substantial global morbidity and mortality each year. Although vaccination is the most effective preventive measure, effectiveness varies and may be limited by dose, adjuvants, and route of administration. Skin-targeted vaccination can leverage abundant antigen-presenting cells and lymphatic networks to enhance immune responses. Microneedle patches (MNPs) offer a reliable, simple, and potentially dose-sparing transdermal delivery platform with logistical advantages over intramuscular (IM) injection. This study investigates whether inactivated influenza vaccination via dissolvable MNPs elicits humoral and cellular immune responses that are comparable to or superior to those induced by IM administration in healthy adults, and explores mechanistic immunologic correlates including HAI, NAI, antibody avidity, cytokine responses, T follicular helper cells, and memory B cells.

Prior work demonstrates that intradermal or transdermal delivery can spare antigen and improve immunogenicity by engaging skin-resident dendritic cells. MNPs have shown multiple advantages in preclinical models, including strong neutralizing antibody responses, cross-reactivity, longer-lasting protection, and dose sparing. A first-in-human phase 1 trial (2015–2016) showed MNPs were safe, well tolerated, preferred by participants, and produced HAI responses at least as strong as IM injection, particularly for influenza B. Beyond HAI, neuraminidase (NA)-specific antibodies are increasingly recognized as independent correlates of protection that can reduce infection severity and shedding. The present study extends earlier findings by examining broader immune mechanisms (NAI, avidity, cytokines/chemokines, Tfh, T cell function, and memory B cells) after MNP versus IM vaccination.

Design: Partially blinded, randomized, placebo-controlled phase 1 study in healthy adults aged 18–49 years. A substudy compared inactivated influenza vaccination (IIV) delivered via dissolvable microneedle patch (MNP; n=11) versus intramuscular (IM; n=11) injection, both administered by healthcare workers. Laboratory analyses were performed blinded. Participants: 22 adults (balanced demographics between groups) from the 2015 trial cohort who received the licensed 2014–2015 trivalent IIV via MNP or IM. Inclusion/exclusion per ClinicalTrials.gov NCT02483432. Vaccine: 2014–2015 trivalent formulation (Seqirus) containing A/Christchurch/16/2010 (H1N1), A/Texas/50/2012 NYMC X-223 (H3N2), and B/Massachusetts/2/2012. The same antigens were used in both formulations. Antigen incorporated into MNPs prepared under GMP; IM administered by hypodermic needle to deltoid; MNP applied to dorsal skin for 20 minutes. Antigen content in used MNPs assessed by SRID; NA content was not measured, and HA-to-NA ratio was assumed similar across formulations. Sampling: Blood collected at Days 0 (pre), 2–3, 8–10 (reported as 10–26 in methods; key cellular timepoints at Day 8), ~Day 28–30, and ~Day 166–194 (6 months). Serum, plasma, and PBMCs processed and stored per standard procedures. Assays:

• HAI: Standard hemagglutination inhibition against the vaccine strains; GMTs, GMFRs, seroprotection, and seroconversion calculated.
• NAI: Neuraminidase inhibition against H6N1, H6N2, H6N3 reassortants; GMTs and fold-changes assessed at Day 28 and Day 180.
• Antibody avidity: Surface plasmon resonance (SPR) with functional rHA1 to determine antibody off-rate constants at Day 28.
• Cytokines/chemokines: 25-plex Luminex panel measured fold changes at Days 2 and 8 post-vaccination relative to baseline.
• Monocytes: Frequency of CD14+ CD16+ monocytes within HLA-DR+ lineage-negative cells over time.
• T cell function: Intracellular cytokine staining of CD4+ T cells stimulated with HA peptide pools; assessment of IL-2, IL-21, IFN-γ, TNF-α, and CD154 (CD40L) at Day 8 relative to baseline.
• Circulating T follicular helper (cTfh) cells: CD4+ CXCR5+ CXCR3− ICOS+ PD-1+ measured at Day 8.
• Memory B cells (MBCs): Influenza-specific IgG-secreting MBCs by ELISpot at baseline and Day 28. Statistics: Student’s t-test and chi-square for demographics; Mann–Whitney/Wilcoxon for between-group comparisons; GMTs with 95% CIs on log-transformed data; Bonferroni correction for multiple comparisons; chi-square for seroconversion/seroresponse rates. Analyses conducted in R 3.6.3 with two-sided alpha 0.05.

• HAI responses: Similar between MNP and IM at 1 and 6 months for H1N1, H3N2, and B strains. HAI increased at Day 28 vs baseline and waned by Day 180 in both groups. GMFR for the B strain was higher with MNP (P=0.009). Seroprotection: 100% at Day 28 for all strains; 82–100% at Day 180 across strains and groups. Seroconversion at Day 28 ranged 55–82% (MNP) vs 18–82% (IM); differences not statistically significant across strains.
• NAI responses: MNP induced significantly greater fold increases in NAI GMTs than IM at Day 28 and Day 180 across H6N1, H6N2, and H6N3 reassortant targets; baseline NAI did not differ between groups.
• Antibody avidity: SPR-measured antibody off-rate constants at Day 28 indicated similar binding avidity between MNP and IM groups.
• Cytokine/chemokine profiles: Early post-vaccination increases with MNP in IP-10, IL-1β, and IL-8 versus baseline, with levels significantly higher than IM; higher IL-5 and IL-13 observed in MNP group, potentially consistent with transient local pruritus noted previously.
• cTfh cells: Significantly higher percentage of cTfh (CD4+ CXCR5+ CXCR3− ICOS+ PD-1+) at Day 8 in MNP versus IM (P=0.04).
• CD4+ T cell function: Trend toward increased IL-2+ CD154+ CD4+ T cells at Day 8 in MNP (P=0.07), but not statistically significant; no significant changes detected for IFN-γ, TNF-α, or IL-21.
• Monocytes: No significant between-group difference in CD14+ CD16+ monocyte frequencies (P=0.33), though a trend toward higher pro-inflammatory monocytes at Days 2–3 was noted with MNP.
• Memory B cells: No significant differences between groups in influenza-specific IgG-secreting MBCs at Day 28 for H1, H3, or N2 antigens.

Targeting the skin with MNPs elicited humoral and cellular responses that were comparable or superior to IM injection, despite a lower antigen dose in MNPs. While HAI—a traditional correlate of protection—was similar between groups, NAI responses were significantly higher and more durable with MNPs, underscoring NA-specific immunity as an independent protective correlate. The MNP route triggered distinct early innate signatures (elevated IP-10, IL-1β, IL-8) and allergic-type cytokines (IL-5, IL-13), as well as a significant increase in circulating T follicular helper cells by Day 8, consistent with enhanced germinal center help and potentially improved antibody quality. However, Day 28 antibody avidity and frequencies of IgG-secreting memory B cells did not differ, and HA-specific CD4+ T cell polyfunctionality showed only a nonsignificant trend. Collectively, these findings support that skin-based delivery via MNPs can enhance select aspects of immunity—particularly NA inhibition and cTfh responses—while maintaining overall serologic protection comparable to IM administration, aligning with proposed mechanisms of improved skin-targeted vaccination.

Dissolvable microneedle patch delivery of trivalent inactivated influenza vaccine in healthy adults generated immune responses that were similar or greater than those elicited by IM injection. Notably, MNPs induced higher neuraminidase-inhibiting antibodies and a significant increase in circulating T follicular helper cells, with comparable HAI titers and antibody avidity to IM vaccination. Given the logistical benefits of MNPs (thermostability potential, simplified administration, acceptability) and the immunologic advantages observed, MNPs represent a promising alternative to IM vaccination for seasonal influenza. Future work should include larger, diverse cohorts; evaluation in immunologically naïve populations (e.g., young children); later timepoints for memory B cell assessment; comprehensive CD8+ T cell analyses; standardized quantification of HA and NA content; and inclusion of placebo MNP immunologic assessments to isolate any patch-related effects.

• Small sample size (n=22 in substudy) limits statistical power and generalizability typical of phase 1 trials.
• Prior influenza vaccinations among participants may have attenuated between-group differences in HAI.
• Potential differences in HA content delivered by MNPs versus IM and lack of NA quantification; HA-to-NA ratios were assumed but not measured in both formulations.
• Memory B cells were measured only at Day 28; later assessments might reveal differences.
• CD8+ T cell responses were not assessed due to limited PBMC availability.
• No extended immunologic analysis of the placebo MNP group, limiting assessment of patch-only effects.
• Young, healthy adult cohort may not represent older adults or high-risk populations.