Effect of an exercise bout before the booster dose of an inactivated SARS-CoV-2 vaccine on immunogenicity in immunocompromised patients

Abstract

This randomized controlled study aimed to investigate whether a single bout of exercise before the homologous booster dose of a SARS-CoV-2 inactivated vaccine could enhance immunogenicity in patients with spondyloarthritis. We selected 60 consecutive patients with spondyloarthritis (SpA). Patients assigned to the intervention group performed an exercise bout comprising three exercises. Then, they remained at rest for 1 h before vaccination. The control group remained at rest before vaccination. Immunogenicity was assessed before (Pre) and 1 mo after (Post) the booster using seropositivity rates of total anti-SARS-CoV-2 S1/S2 IgG, geometric mean titers of anti-S1/S2 IgG (GMT), frequency of neutralizing antibodies (NAb) positivity, and NAb activity. At Pre, 16 patients from the exercise group and 16 patients from the control group exhibited seropositivity for IgG (59% vs. 57.1%), and 1 mo after the booster dose, seropositivity occurred in 96% versus 100% of the cases. Only 10 patients from the exercise group and 12 patients from the control group showed positive NAb serology at Pre (37% vs. 42.8%). One month following the booster, NAb positivity was 96% versus 93%. GMT was comparable between groups at Pre. At Post, GMT increased similarly in both groups. Likewise, NAb activity was similar between groups at Pre and increased similarly in both of them as a result of the booster (47.5% vs. 39.9%). In conclusion, a single bout of exercise did not enhance immunogenicity to a homologous booster dose of an inactivated SARS-CoV-2 vaccine among patients with spondyloarthritis.

NEW & NOTEWORTHY We tested the role of exercise as an adjuvant to a booster of a COVID-19 vaccine. Immunocompromised patients were immunized after an acute bout of exercise or not. Patients exhibited an excellent immunogenicity in response to the booster dose. Exercise did not add to the vaccine effects on IgG or neutralizing antibodies.

INTRODUCTION

Vaccination is a key strategy to mitigate mortality and morbidity rates for COVID-19. In countries with high capacity of rapid rollouts, deaths have been dramatically reduced, even in the presence of the more contagious variants (1). However, vaccine responses largely vary between individuals. This is the case of immunocompromised individuals who generally exhibit lower immunogenicity to SARS-CoV vaccines, likely due to the disease-related immune dysfunction and the use immunosuppressive, glucocorticoid, and/or biological therapies (2–5). Prolonged SARS-CoV-2 shedding, decreased SARS-CoV-2 clearance, and enhanced viral genomic evolution have been reported in immunocompromised patients, which underscores the relevance of improving vaccine responses in this population to ultimately prevent the emergence of new variants (6, 7).

Regular exercise can enhance the responses to some vaccines (e.g., influenza) in individuals with dysfunctional immune system (8–10), an observation recently extended to a SARS-CoV-2 vaccine (11). Of relevance, a single exercise session has been shown to be sufficient to enhance the humoral and cellular responses to vaccines (12, 13). This immunoenhancement effect seems to be associated with the exercise-induced local inflammatory process, resulting in immediate secretion of cytokines (e.g., IL-6, IL-7, and IL-15), edema, hyperemia, increased vascular permeability, enhanced interferon-γ responses, increased neutralizing antibodies and cytotoxic T cell number, and activity (13–15). However, to our knowledge, this hypothesis has not yet been tested in the context of SARS-CoV vaccines.

In this exploratory study, within a phase 4 vaccination trial of patients with autoimmune rheumatic diseases (4), we tested whether a single bout of exercise before the homologous booster dose of a SARS-CoV-2 inactivated vaccine, Sinovac-CoronaVac, the most frequently administered vaccine worldwide (16), could enhance immunogenicity in patients with spondyloarthritis (SpA), who were shown to experience suboptimal vaccine responses following a two-dose vaccination course (4).

METHODS

Participants

This was a randomized controlled trial (1:1) within a single-arm, phase 4 vaccination trial, conducted in a tertiary hospital in Sao Paulo, Brazil. The main trial was designed to test the effectiveness and safety of Sinovac-CoronaVac in a large cohort of patients with systemic autoimmune rheumatic diseases (clinicaltrials.gov No. NCT04754698). This substudy was approved by the local ethics committee as an addendum to the main trial (CAAE: 42566621.0.0000.0068), and all patients agreed to participate by providing an informed written consent. On the basis of feasibility, resources, capacity of research staff and facility, and patients’ availability (17, 18), we selected 60 consecutive patients with SpA (ankylosing spondylitis and psoriatic arthritis), diagnosed according to established criteria (19). These patients had been invited to receive the homologous booster dose of Sinovac-CoronaVac (6 mo after the second dose) in the context of the main trial. Patients with SpA (ankylosing spondylitis and psoriatic arthritis) were eligible. Age <50 years, current use of methotrexate, and any health condition precluding exercise were considered exclusion criteria. The randomization list for intervention (exercise) or control (no exercise) was created using a computer-generated code.

Experimental Design

Patients assigned to the exercise group performed an exercise bout in our intrahospital exercise physiology laboratory, located at the same medical center where vaccination was administered. The exercise session was conducted by experienced trainers and was composed of a specific warm-up followed by three unilateral strength exercises using dumbbells (i.e., scapular abduction, elbow flexion, and elbow extension) involving eccentric and concentric contractions. Four sets of 8–12 maximum repetitions (until concentric failure) were performed for each exercise, interspaced by a 30-s interval. The total duration of the exercise session was ∼20 min. After exercising, patients remained at rest for 1 h before vaccination (third dose of CoronaVac – Sinovac Life Sciences, Beijing, China), injected into the exercised arm. The control group was required to remain seated for 1 h before vaccination. All patients underwent either the exercise or control conditions and were vaccinated on a single day (September 18, 2021). As a proxy of exercise-induced local edema/inflammation, maximal circumference of the biceps at elbow flexion was measured before and after exercise using a measuring tape. In addition, muscle soreness of the exercised limb was assessed before and after 0, 24, and 48 h from the exercise session, using a visual analog scale ranging from 0 (no pain) to 10 (worst possible pain).

Immunogenicity Outcomes

Blood samples (20 mL) were collected immediately before (Pre) and ∼1 mo after the third dose injection (Post) and were stored in a freezer at −70°C. Laboratory staff was blinded to the intervention.

Anti-SARS-CoV-2 S1/S2 IgG antibodies.

Human IgG antibodies against S1 and S2 proteins in RBD (Indirect ELISA, LIAISON SARS-CoV-2 S1/S2 IgG, DiaSorin, Italy) were assessed by chemiluminescent immunoassay. SC was defined as positive serology (>15.0 UA/mL) after vaccination (4, 11).

SARS-CoV-2 cPass virus-NAb.

Circulating neutralizing antibodies (NAb) against SARS-CoV-2 were assessed using the SARS-CoV-2 sVNT Kit (GenScript, Piscataway, NJ), which detects neutralizing antibodies that block the interaction between RBD in the viral spike glycoprotein with angiotensin-converting enzyme 2 (ACE2) cell surface receptor. Tests were performed on ETI-MAX-3000 (DiaSorin, Italy). Samples were classified as either “positive” or “negative” (inhibition ≥ 30 or < 30%, respectively), as suggested by the manufacturer (4, 11).

Physical Activity Level

Using a telephone-based survey, physical activity was assessed 1 mo before the first vaccine dose in four domains: leisure time, household activities, work, and commuting. Participants were classified as physically active or inactive according to World Health Organization (WHO) guidelines (i.e., physical inactivity defined as <150 min/wk of moderate-to-vigorous intensity aerobic activity) (20).

Statistical Analysis

Generalized estimating equations (GEE) models were used to assess possible differences between groups, assuming fixed effects for group and time and random effects for patients. Post hoc tests with Tukey’s adjustment were performed for multiple pairwise comparisons purposes when necessary. Two exploratory analyses were conducted. First, sex, disease (ankylosing spondylitis and psoriatic arthritis), and age were used as covariates in GEE models. Second, GEE models were also used to assess possible differences between groups for GMT and NAb activity in an exploratory, sensitivity analysis involving only patients who did not exhibit seropositivity for IgG (exercise: n = 11; control: n = 12) or NAb (exercise: n = 17; control: n = 16) before the booster dose. Data are expressed as absolute and relative frequencies or estimated mean difference between groups at Post (EMD) ± 95% confidence intervals (95%CI), and relative delta change (%), excepted otherwise stated. Significance level was previously set at P ≤ 0.05. Analyses were performed in RStudio version 4.02, using the following packages: “geepack,” “stats,” “tydverse,” “emmeans,” “ggplot2,” and “ggpubr.”

RESULTS

Sixty patients were randomly assigned to either exercise or control group (n = 30/group). Three patients in the exercise group did not receive the booster dose (due to personal reasons), whereas two patients in the control group did not attend the postvaccination visit to collect blood sample. Therefore, 27 patients in the exercise group and 28 patients in the control group were included in the analyses (Fig. 1). Patients’ main characteristics are described in Table 1.

Figure 1.

Flowchart of participants.

Table 1.

Characteristics of patients in the exercise and control groups

	Exercise	Control
n	27	28
Age, yr (means ± SD)	58.8 ± 5.3	60.8 ± 5.8
Body mass index, kg/m² (mean ± SD)	28.0 ± 8.0	27.1 ± 4.4
Disease duration, yr (mean ± SD)	28.6 ± 10.7	28.5 ± 11.0
Sex, n (Male/Female)	20/7	20/8
Disease, n (%)
Ankylosing spondylitis	20 (74)	20 (71.4)
Psoriatic arthritis	7 (26)	8 (28.6)
Treatment, n (%)
Biological therapy*	10 (37.0)	19 (67.8)
Immunosuppressants	4 (15.0)	4 (14.2)
Prednisone	4 (15.0)	1 (3.5)
Comorbidities, n (%)
Arterial hypertension	20 (74.0)	14 (50.0)
Diabetes mellitus	5 (19.0)	11 (39.2)
Chronic renal disease	2 (7.0)	2 (7.1)
IgG positivity, n (%)	16 (59.0)	16 (57.1)
NAb positivity, n (%)	10 (37.0)	12 (42.8)
Physically inactive patients, n (%)Ϯ	14 (51.8)	10 (37.0)

*Biological therapy: TNF inhibitor (exercise n = 4; control n = 16), monoclonal antibodies (exercise n = 10; control n = 17), etanercept (exercise n = 0; control n = 4), adalimumab (exercise n = 3; control n = 7), infliximab (exercise n = 1; control n = 5), certolizumab pegol (exercise n = 1; control n = 0); immunosuppressants: leflunomide (exercise n = 4; control n = 3), cyclosporin (exercise n = 1; control n = 0). Sums to more than the patient numbers provided because some patients were taking more than one biological agent.

ϮOne missing data in the control group.

Before the exercise intervention, mean relaxed biceps circumference was 31.57 ± 4.09 and mean contracted biceps circumference was 32.42 ± 3.99. After exercise, relaxed and contracted biceps circumferences were 32.67 ± 4.19 and 33.77 ± 4.12, respectively (all P < 0.01 when compared with before exercise). Likewise, muscle soreness was also increased across time (before: 1.1 ± 2.1; immediately after exercise: 2.6 ± 2.8; 24 h: 4.4 ± 2.5; 48 h: 3.5 ± 2.5; all P < 0.01 vs. preexercise). These data suggest that exercise effectively induced muscle edema/inflammation.

At Pre, 16 patients from the exercise group and 16 patients from the control group exhibited similar seropositivity rates for IgG (59% vs. 57.1%, respectively). One month after the third dose, IgG seropositivity rates were comparable between groups (96% vs. 100%), with no group-by-time interaction (P = 0.84).

As for NAb positivity, 10 patients from the exercise group and 12 patients from the control group showed positive serology (37% vs. 42.8%) at Pre. One month following the booster dose, positive serology was similar between the two groups (96% vs. 93%), with no group-by-time interaction (P = 0.41).

GMT was comparable between exercise and control groups at Pre (P > 0.05). After the booster dose, GMT increased similarly in both groups (56.9% vs. 57.9%), with no group-by-time interaction (P = 0.82; EMD: −40.4 UA/mL, 95%CI: −327, 246 UA/mL). Likewise, NAb activity was similar between groups at Pre, and it increased similarly in both of them after the booster (47.5% vs. 39.9%); there was no group-by-time interaction (P = 0.99; EMD: −6.19%, 95%CI: −17; 4.6%) (Fig. 2). Models adjusted by sex, disease (ankylosing spondylitis and psoriatic arthritis), and age as covariates yielded similar results (data not shown).

Figure 2.

A: IgG positivity rates were comparable between groups before (Pre) and increased similarly in both of them after the booster (group-by-time interaction, P = 0.84). B: NAb positivity rates were also comparable between groups at Pre, and it increased similarly in both of them after the booster (group-by-time interaction, P = 0.41). C: geometric mean titers of anti-S1/S2 IgG (GMT) were comparable between groups at Pre, and increased similarly in both of them as a result of the booster (exercise: 56.9%; control: 57.9%), with no group-by-time interaction (EMD: −40.4 UA/mL, 95%CI: −327, 246 UA/mL, P = 0.82). D: NAb activity was similar between groups at Pre and increased similarly in both of them after the booster (exercise: 47.5%, control: 39.9%), with no group-by-time interaction (EMD: −6.19%, 95%CI: −17; 4.6%, P = 0.99). Data were analyzed using GEE. Boxplot are expressed as median ± interquartile (minimum; maximum). GEE, generalized estimating equations; NAb, neutralizing antibodies.

In the sensitivity analysis involving only patients who exhibited no seropositivity for IgG and NAb at Pre, similar increases in GMT (82.2% vs. 92.6%) and NAb activity (66.3% vs. 66.5%) were observed in both groups following the booster (main effect of time, P < 0.01 for both), with no group-by-time interactions (P = 0.24 and P = 0.50 for GMT and NAb activity; EMD: −110 UA/mL, 95%CI: −292, 72.5 UA/mL and EMD: −3.1%, 95%CI: −12.2, 5.9%, respectively).

DISCUSSION

The search for strategies able to enhance SARS-CoV-2 vaccine responses is key to countering the pandemic, particularly for individuals with dysfunctional immune system. Growing evidence now shows that immunocompromised individuals respond less robustly to different SARS-CoV-2 vaccines (4). This is the case for patients with SpA, which is immune-mediated inflammatory disease characterized by axial and/or peripheral joint inflammation, as well as extra-articular manifestations. In the previous phase of this phase 4 trial, a two-dose course of Sinovac-CorovaVac (4), a WHO-approved vaccine that protects against severe cases and mortality of COVID-19 (21), led to a moderate immunogenicity among immunocompromised patients (including those with SpA). In this substudy, within the phase 4 trial, we speculated that an acute exercise session could enhance the responses of the booster dose in this group, a hypothesis that was refuted by the current findings.

There is evidence that physical activity improves immune response to vaccination, particularly in individuals at risk for immune dysfunction (11). For instance, studies involving older adults show increased responses to vaccination in physically active individuals versus inactive ones, or in those who received an exercise intervention versus nonexercised ones (8, 22, 23). However, no studies so far have assessed the effects of exercise on SARS-CoV-2 vaccines. Recently, we also showed that among a large cohort of patients with systemic autoimmune rheumatic diseases, those who were physically active exhibited higher levels of antibodies following two doses of Sinovac-CoronaVac when compared with inactive individuals (11). Next, we tested the immunoenhancement effect of an exercise session in a subgroup of this cohort (i.e., SpA) on the basis of evidence showing that a single bout of exercise at the time of vaccination can elicit better immune responses (12). The reason underlying the null findings of the present study may rely in the fact that the immunoenhancement effect of exercise commonly manifests for antigens eliciting weak immune responses (12, 23, 24). This notion is well illustrated by the findings showing that exercise was able to enhance the responses to a reduced dose of pneumococcal vaccine, but did not change the responses to the full dose (12). In addition, a 45-min walk was shown to be immunostimulatory in older women who had lower prevaccine titers, but not in older men with stronger immune responses (24).

To test the hypothesis that patients with weaker immunity would primarily benefit from the immunoenhancement effect of exercise, we subanalyzed the patients with no NAb or IgG seropositivity before the booster. The results indicated no effect of the intervention, corroborating the lack of immunostimulatory effects of exercise in this study. It is likely that the booster shot itself resulted in such a robust antibody response that masked any adjuvant role of exercise. In fact, 1 mo following the booster dose, out of the 55 patients analyzed, only 1 did not exhibit IgG seropositivity, and 3 did not show NAb positivity. This excellent immunogenicity to the booster may have created a ceiling effect for exercise.

This study has limitations. First, the sample size was relatively low, despite the homogenous characteristics of the participants. Second, despite the fact that this study investigated diseases that represent the archetype of immune system dysfunction, the current findings cannot be promptly generalized to other conditions affecting the immune function. Third, as previously discussed, this three-dose schedule resulted in a strong immune response, which may have masked any enhancement effect of exercise. Further studies assessing exercise as an adjuvant to first, second, or even booster half doses remain necessary. Fourth, data from our targeted exercise model (i.e., focused on muscle around the injection site) may not be generalized to other exercise models (e.g., whole body resistance exercises, running, walking, etc.). Fifth, the present data are confined to timing of the exercise relative to vaccination employed in this study, although previous data showed no effect of vaccine timing on the efficacy of exercise intervention on the immune responses to an influenza vaccine (25). Finally, regular exercise may confer enhanced vaccine responses through distinct mechanisms from those of acute exercise [for a review, see Pascoe et al., (23)], so that the present findings are restricted to the latter model of intervention.

In conclusion, a single bout of exercise did not enhance the excellent immunogenicity in response to a homologous booster dose of an inactivated SARS-CoV-2 vaccine among immunocompromised patients with SpA.

GRANTS

This trial is sponsored by grants from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) (#2015/03756-4 to N.E.A., S.G.P., C.A.S. and E.B.; #2015/26937-4 to A.J.P.; #2017/13552-2 to B.G.; #2020/04877-8 to I.R.L.; #2017/23688-9 to R.P.S.; #2019/14820-6 to B.C.M.; #2019/14819-8 to F.I.S.; #2019/15231-4 to S.M.S.), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (#305242/2019-9 to E.B., #304984/2020-5 to C.A.S.) and B3-Bolsa de Valores do Brasil.

DISCLAIMER

Instituto Butantan supplied the study’s product and had no other role in the trial.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

B.G., C.G.S.S., S.M.S., A.C.M-R., C.A.S., R.M.R.P., S.K.S., D.C.O.A., P.D.S-B., H.R., and E.B. conceived and designed research; B.G., S.M.S., I.R.L., R.P.d.S., A.J.P., B.C.M., F.I.S., S.G., G.O-J., and E.B. performed experiments; B.G., S.M.S., R.P.d.S., and A.J.P. analyzed data; B.G., C.G.S.S., S.M.S., R.P.d.S., A.J.P., P.D.S-B., and H.R. interpreted results of experiments; B.G. and S.M.S. prepared figures; B.G., C.G.S.S., P.D.S-B., H.R., and E.B. drafted manuscript; B.G., C.G.S.S., S.M.S., I.R.L., R.P.d.S., A.J.P., B.C.M., F.I.S., S.G., G.O-J., N.E.A., A.C.M-R., C.A.S., E.F.N.Y., S.G.P., R.M.R.P., S.K.S., D.C.O.A., P.D.S-B., H.R., and E.B. edited and revised manuscript; B.G., C.G.S.S., S.M.S., I.R.L., R.P.d.S., A.J.P., B.C.M., F.I.S., S.G., G.O-J., N.E.A., A.C.M-R., C.A.S., E.F.N.Y., S.G.P., R.M.R.P., S.K.S., D.C.O.A., P.D.S-B., H.R., and E.B. approved final version of manuscript.