The effect of metformin on influenza vaccine responses in nondiabetic older adults: a pilot trial

Abstract

Background: Aging is associated with progressive declines in immune responses leading to increased risk of severe infection and diminished vaccination responses. Influenza (flu) is a leading killer of older adults despite availability of seasonal vaccines. Geroscience-guided interventions targeting biological aging could offer transformational approaches to reverse broad declines in immune responses with aging. Here, we evaluated effects of metformin, an FDA approved diabetes drug and candidate anti-aging drug, on flu vaccination responses and markers of immunological resilience in a pilot and feasibility double-blinded placebo-controlled study.

Results: Healthy older adults (non-diabetic/non-prediabetic, age: 74.4 ± 1.7 years) were randomized to metformin (n = 8, 1500 mg extended release/daily) or placebo (n = 7) treatment for 20 weeks and were vaccinated with high-dose flu vaccine after 10 weeks of treatment. Peripheral blood mononuclear cells (PBMCs), serum, and plasma were collected prior to treatment, immediately prior to vaccination, and 1, 5, and 10 weeks post vaccination. Increased serum antibody titers were observed post vaccination with no significant differences between groups. Metformin treatment led to trending increases in circulating T follicular helper cells post-vaccination. Furthermore, 20 weeks of metformin treatment reduced expression of exhaustion marker CD57 in circulating CD4 T cells.

Conclusions: Pre-vaccination metformin treatment improved some components of flu vaccine responses and reduced some markers of T cell exhaustion without serious adverse events in nondiabetic older adults. Thus, our findings highlight the potential utility of metformin to improve flu vaccine responses and reduce age-related immune exhaustion in older adults, providing improved immunological resilience in nondiabetic older adults.

Keywords: Aging; Geroscience-guided clinical trial; Immune exhaustion; Metformin; Vaccination.

Fig. 1

Experimental design and assessment of serum antibody titers. A) Experimental design. Patients were randomized to metformin (final dose 1500 mg ER/day) or placebo and started with 1 tablet/day for week 1 (500 mg metformin ER/day or placebo), then 2 tablets a day for week 2 (1000 mg metformin ER or placebo), and finally 3 tablets a day for week 3 (1500 mg metformin ER or placebo) until the completion of the study. Participants were vaccinated with high-dose flu vaccine after ~ 10 weeks of treatment. Blood was drawn prior to treatment (week -10, Baseline), immediately prior to vaccination (week 0), and 1-, 5-, and 10-weeks post vaccination. Serum was analyzed via hemagglutinin inhibition assay for flu antibody titers. Geometric Mean Titers (GMT) were normalized to pre-vaccination (Pre-Vax) levels to calculate fold change for the B) A/Kansas/14/2017 (H3N2)-like virus, C) A/Brisbane/02/2018 (H1N1)pdm09-like virus and D) B/Colorado/06/2017-like virus (B/Victoria lineage) contained in the seasonal flu vaccine. The dashed line denotes a fold change of 1. Statistical significance was calculated by two-way repeated measures ANOVA with Šídák’s posthoc corrections and significance set at p < 0.05

Fig. 2

Assessment of ex vivo cell-mediated vaccination responses. PBMCs were stimulated with live flu virus and cell-mediated responses were analyzed in the supernatants and lysates. Inducible Granzyme B (iGrB) levels were calculated in lysates by normalizing to total protein content and unstimulated normalized GrB levels. iGrB and IFNy/IL-10 ratio are presented as fold change at given week post vaccination compared to pre-vaccination levels. No differences were observed in iGrB in response to influenza A) AVic/H3N2 or B) BLee strain or in IFNy/IL-10 responses to C) Avic/H3N2 or D) Blee strains. Statistical significance was calculated by two-way repeated measures ANOVA with Šídák’s posthoc corrections and significance set at p < 0.05

Fig. 3

Flow cytometry analysis of circulating PBMC responses to flu vaccination. Peripheral blood mononuclear cells (PBMCs) were analyzed for A) circulating T follicular helper cells (cTfh), B) activated ICOS + PD1 + cTfh, and C) Antibody Secreting Cells (ASCs). Data are presented as fold change at given week post vaccination compared to pre-vaccination levels. A trending main metformin treatment effect was observed in cTfh. Statistical significance was calculated by two-way repeated measures ANOVA with Šídák’s posthoc corrections and significance set at p < 0.05

Fig. 4

Flow cytometry analysis of exhaustion markers. CD4 and CD8 T cells were analyzed for expression of exhaustion markers CD57 and PD-1 prior to and following 20 weeks of treatment with either placebo or metformin and fold change in mean fluorescent intensity (MFI) was calculated. CD4 T cells had reduced A) CD57 expression with metformin, while no differences in C) PD-1 expression were observed. CD8 T cells had no differences in B) CD57 expression, while there was a trending increased in D) PD-1 expression with metformin treatment. Statistical significance was calculated by t-test and significance set at p < 0.05

Fig. 5

Analysis of circulating markers of inflammation and aging. Circulating markers of inflammation, growth factors, and age-related factors were measured prior to and following 20 weeks of treatment with either placebo or metformin. Plasma A) IL-6, B) IL-8, C) IL-17A, D) TNFa, E) TNFRI, F) TNFRII, G) C-reactive protein, H) IGF-1, I) IGFBP-2, J) Osteopontin, and M) GDF-15 and serum K) MMP-2 and L) MMP-9 was measured and fold change was calculated. Statistical significance was calculated by t-test and significance set at p < 0.05