Studies of cell-mediated immune responses to influenza vaccination in systemic lupus erythematosus

Authors


Abstract

Objective

Both antibody and cell-mediated responses are involved in the defense against influenza. In patients with systemic lupus erythematosus (SLE), a decreased antibody response to subunit influenza vaccine has been demonstrated, but cell-mediated responses have not yet been assessed. This study was therefore undertaken to assess cell-mediated responses to influenza vaccination in patients with SLE.

Methods

Fifty-four patients with SLE and 54 healthy control subjects received subunit influenza vaccine. Peripheral blood mononuclear cells and sera were obtained before and 1 month after vaccination. Cell-mediated responses to A/H1N1 and A/H3N2 vaccines were evaluated using an interferon-γ (IFNγ) enzyme-linked immunospot assay and flow cytometry. Antibody responses were measured using a hemagglutination inhibition test.

Results

Prior to vaccination, patients with SLE had fewer IFNγ spot-forming cells against A/H1N1 compared with control subjects and a lower frequency of IFNγ-positive CD8+ T cells. After vaccination, the number of IFNγ spot-forming cells increased in both patients and control subjects, although the number remained lower in patients. In addition, the frequencies of CD4+ T cells producing tumor necrosis factor and interleukin-2 were lower in patients after vaccination compared with healthy control subjects. As expected for a subunit vaccine, vaccination did not induce a CD8+ T cell response. For A/H3N2-specific responses, results were comparable. Diminished cell-mediated responses to influenza vaccination were associated with the use of prednisone and/or azathioprine. The increase in A/H1N1-specific and A/H3N2-specific antibody titers after vaccination was lower in patients compared with control subjects.

Conclusion

In addition to a decreased antibody response, cell-mediated responses to influenza vaccination are diminished in patients with SLE, which may reflect the effects of the concomitant use of immunosuppressive drugs. This may render these patients more susceptible to (complicated) influenza infections.

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease characterized by a remitting and relapsing course. Patients with SLE have an increased risk of infection, due to both intrinsic disturbances of immune responses and treatment with immunosuppressive drugs, which is often needed to control disease activity. Indeed, infection-related morbidity and mortality are more frequent in patients with SLE (1).

Influenza infection–related morbidity and mortality are increased in immunocompromised patients (2). Because the incidence of influenza infection is high, with an estimated 5–20% of the general population infected annually (3), influenza vaccination is a clinically relevant issue in patients with SLE. Influenza vaccination of patients with SLE is safe, as it has been shown that influenza vaccination does not induce disease activity (4). Annual vaccination in patients with SLE is therefore recommended (5).

In the immune response to influenza, both antibody and cell-mediated responses, comprising production of CD4+ and CD8+ T cells, are involved. In SLE, antibody responses to influenza vaccination are diminished (6), but cell-mediated responses have not been assessed. The latter are relevant, because it has been shown that in certain groups, such as the elderly, cell-mediated responses to influenza vaccination can be a marker of clinical protection, independent from antibody responses (7). The most frequently used vaccine formulations are split virus or subunit vaccines. With these vaccines, antigens are primarily presented via class II major histocompatibility complex (MHC), which induces CD4+ T cell stimulation (8). However, they are incapable of inducing class I MHC–restricted CD8+ T cell responses (9). In addition, subunit vaccines, in contrast to split virus and whole virus vaccines, do not contain any of the internal proteins that may more readily (re)activate influenza-specific CD8+ T cells.

In SLE, decreased T helper cell recall responses to influenza A and tetanus toxoid antigens have been reported in a subset of patients, as measured by interleukin-2 (IL-2) production upon stimulation. This decreased function could not be accounted for by the use of immunosuppressive agents alone and was shown to be associated with disease activity (10). In addition, lower levels of cell-mediated cytotoxicity against target cells infected with influenza A and influenza B have been observed in patients with SLE (11).

Based on these data, we hypothesized that patients with SLE have lower CD4+ T cell responses to subunit influenza vaccine and lower CD8+ T cell recall responses to influenza antigens compared with healthy control subjects. Cell-mediated responses against influenza in SLE, prior to and following vaccination, were evaluated. In addition, antibody responses were evaluated, and data regarding vaccine safety were recorded.

PATIENTS AND METHODS

Study population.

Patients who were eligible for the study fulfilled at least 4 of the American College of Rheumatology criteria for SLE (12). Exclusion criteria were pregnancy and the presence of an indication for yearly influenza vaccination based on concomitant disease according to international guidelines (13). Healthy individuals who were age- and sex-matched to the vaccinated patients with SLE were included as the control group. Pregnancy was an exclusion criterion for participation as a control subject.

Study design.

Patients with SLE and control subjects were included from October 2005 to December 2005. Before entry, patients were randomized (2:1) to receive an influenza vaccination or to serve as a nonvaccinated patient control. At entry (visit 1), patients randomized for vaccination and all healthy control subjects were vaccinated. Patients and control subjects were followed up at 28 days (visit 2) and 3–4 months after inclusion (visit 3). Peripheral blood mononuclear cells (PBMCs) were isolated from vaccinated participants at visits 1 and 2 (see below). At each visit, blood was drawn, and serum was stored at −20°C until used. Also, the SLE Disease Activity Index (SLEDAI) (14) score was recorded, and patients were asked to mark a 0–10-cm visual analog scale (VAS) for disease activity, where 0 = no activity and 10 = highest activity. Information on influenza vaccination in the previous year was obtained. Adverse reactions to vaccination were recorded using a standardized questionnaire that included the following: itching, pain, erythema, induration at the site of vaccination, shivers, myalgia, fever, headache, nausea, arthralgia, diarrhea, and use of an analgesic/antiinflammatory drug. The study was approved by the institutional medical ethics committee, and informed consent was obtained from all participants.

Influenza vaccine.

A single dose of a trivalent subunit influenza vaccine (Influvac, 2005–2006; Solvay Pharmaceuticals, Weesp, The Netherlands) containing A/New Caledonia/20/99 (AH1N1), A/NewYork/55/2004 (AH3N2), and B/Hong Kong/330/2001 was administered intramuscularly.

Isolation, storage, and thawing of PBMCs.

PBMCs were isolated from heparinized venous blood by density-gradient centrifugation on Lymphoprep (Axis-Shield, Oslo, Norway) immediately after blood was drawn. PBMCs were frozen in RPMI 1640 (Cambrex BioScience, Verviers, Belgium) supplemented with 10% fetal calf serum (FCS), 50 μg/ml of gentamicin (Gibco, Paisley, UK), and 10% dimethylsulfoxide. PBMCs were stored in liquid nitrogen until used. Prevaccination and postvaccination samples from a patient with SLE and a matched control subject were simultaneously thawed and batch-processed. A minimum cell viability of >90%, as evaluated by trypan blue staining, was required. Preceding the performance of the enzyme-linked immunospot (ELISpot) assays, PBMCs were rested by overnight incubation at 37°C. Cells were counted before plating, using an automated cell counter (Beckman Coulter, Fullerton, CA).

Influenza antigens used in assays of cell-mediated responses.

β-propiolactone–inactivated whole virus (WIV) of A/H1N1 and A/Hiroshima/52/2005 (A/H3N2) were used to stimulate PBMCs. A/Hiroshima/52/2005 is a very closely related antigenic variant of the vaccine strain A/NewYork/55/2004.

Interferon-γ (IFNγ) ELISpot assay.

Nitrocellulose plates (Nunc, Rochester, NY) were coated overnight at 4°C with 50 μl of anti-human IFNγ, 15 μg/ml per well (Mabtech, Nacka Strand, Sweden). Plates were washed and blocked with culture medium (RPMI supplemented with 50 μg/ml of gentamicin and 10% FCS) for 1 hour at room temperature. Subsequently, 2 × 105 PBMCs were added per well, in 200 μl, and incubated in culture medium at 37°C with WIV of A/H1N1 and A/H3N2, at a final concentration of 5 μg total viral protein/ml. Stimulation with concanavalin A (5 μg/ml) was used as a positive control, and a negative control consisted of PBMCs in culture medium alone. Stimulation tests were performed in duplicate. After 48 hours, plates were washed with phosphate buffered saline (PBS), and 50 μl of 1 μg/ml biotinylated anti-human IFNγ (Mabtech) was added per well for 3 hours at room temperature. Next, plates were washed again, and 50 μl of streptavidin–alkaline phosphatase (1:1,000; Mabtech) per well was added for 1.5 hours at room temperature. Plates were washed, and 100 μl of BCIP/NBTplus substrate (Mabtech) was added per well for 10 minutes. Finally, plates were washed with tap water. After drying, spots were counted using an automated reader (automated ELISpot video analysis system; Sanquin, Amsterdam, The Netherlands). Results are referred to as IFNγ spot-forming cells, because generation of IFNγ-producing CD4+ and CD8+ T cells as well as natural killer (NK) cells following WIV stimulation have been described (15).

Flow cytometry.

For stimulations, 1.0–1.5 × 106 PBMCs were cultured in 200 μl of culture medium, in 5-ml polypropylene round-bottomed Falcon tubes (Becton Dickinson, Franklin Lakes, NJ). Staphylococcal enterotoxin B (SEB) (Sigma-Aldrich, St. Louis, MO) at 5 μg/ml was used as a positive control. WIV A/H1N1 and WIV A/H3N2 were used at final concentrations of 1 μg of total viral protein/ml. WIV and negative control (medium only) cultures were incubated in the presence of 10 μg/ml anti-CD28/CD49 (Becton Dickinson). Cells were incubated for 18 hours at 37°C; for the final 16 hours, cells were incubated in the presence of 10 μg/ml brefeldin A (Sigma-Aldrich). Following incubation, 10 μl of 40 mM EDTA in PBS was added, and tubes were vortexed and incubated for 10 minutes to facilitate resuspending. Next, 2 ml of fluorescence-activated cell sorting lysing solution (Becton Dickinson) was added for 10 minutes. Cells were spun down and washed in PBS/1% bovine serum albumin.

Subsequently, cells were permeabilized in 500 μl Perm II (Becton Dickinson) for 10 minutes in the dark in the presence of Pacific Blue and Pacific Orange dyes (Invitrogen, Carlsbad, CA), in a different combination for each stimulus, to enable fluorescent cell bar coding (16). PBS/20% FCS was added for 5 minutes. Cells were washed and pooled per PBMC sample. Next, fluorescein isothiocyanate–conjugated anti-CD3, phycoerythrin–Cy7 (PE-Cy7)–conjugated anti-CD4, peridinin chlorophyll A protein–conjugated anti-CD8, allophycocyanin (APC)–Cy7–conjugated anti-CD69, Alexa 700–conjugated anti-IFNγ, APC-conjugated anti–tumor necrosis factor (anti-TNF), and PE-conjugated anti–interleukin-2 (anti–IL-2) (all from Becton Dickinson) were added, according to the manufacturer's instructions. After incubation for 30 minutes at room temperature, cells were washed and immediately analyzed on a LSR II flow cytometer (Becton Dickinson). Data for at least 1 × 106 CD3+ cells were collected.

Using the WinList software package (Verity Software House, Topsham, ME), positively and negatively stained populations were gated, and Boolean gating was applied. First, lymphocytes were gated by CD3 expression and sideward scatter patterns. Next, CD4+ and CD8+ T cell populations were gated as CD4+CD8− or CD4−CD8+, respectively. Then, cells from different stimulation tubes were separated in a Pacific Blue/Pacific Orange dye plot. Finally CD69+/−, cytokine+/− quadrants were set for the different stimuli simultaneously, according to the negative and positive controls. Percentages of antigen-specific cells were expressed as the percentage of CD69+ cytokine-producing CD4+ or CD8+ T cells within the total CD4+ or CD8+ T cell population.

Antibody response to influenza.

For quantitative detection of anti-influenza antibodies, the hemagglutination inhibition test was employed, following standard procedures (17). Influenza A/H1N1 and A/H3N2 vaccines were provided by Solvay Pharmaceuticals. Seroconversions were defined as a 4-fold rise in titer 1 month after vaccination, and seroprotection was defined as a titer ≥40. For calculation purposes, titers <10 (below the level of detection) were assigned a value of 5 (18).

Statistical analysis.

Data were analyzed using SPSS version 14 software (SPSS, Chicago, IL). Titers were log-transformed prior to testing of the geometric mean titers (GMTs). For comparisons of T cell cytokine responses, the Mann-Whitney U test and Wilcoxon's signed rank test were used. All T cell frequencies are reported after background subtraction of the frequency of the identically gated population of cells from the same sample stimulated without antigen. For correlations, Spearman's rho was used. Age was normally distributed and tested with Student's t-test. For all other variables, Fisher's exact test and the Mann-Whitney U test were used, where appropriate. Two-sided P values less than 0.05 were considered significant. No adjustments for multiple testing were made, given the exploratory design of the study.

RESULTS

Patient characteristics.

Eighty patients with SLE gave informed consent to participate and were randomized (54 to the vaccination group and 26 to the nonvaccination group). Two patients initially randomized to the nonvaccination group were excluded (due to pregnancy and study withdrawal, respectively). Patient groups did not differ in sex, age, and medication use. More patients in the vaccination group had received an influenza vaccination the previous year compared with patients in the nonvaccination group and control subjects (Table 1).

Table 1. Baseline characteristics and disease parameters in patients with SLE and healthy control subjects*
 SLE patientsControls, vaccinated (n = 54)
Nonvaccinated (n = 24)Vaccinated (n = 54)
  • *

    Except where indicated otherwise, values are the number (%). SLE = systemic lupus erythematosus; NA = not applicable; SLEDAI = Systemic Lupus Erythematosus Disease Activity Index; VAS = visual analog scale.

  • P < 0.05 versus nonvaccinated patients.

  • P < 0.001 versus vaccinated patients.

  • §

    Five patients received 15 mg/week, and 1 patient received 25 mg/week.

Male sex2 (8.3)10 (18.5)11 (20.4)
Age, mean ± SD years45.5 ± 11.544.8 ± 13.643.1 ± 10.9
Influenza vaccination in previous year9 (37.5)34 (63.0)3 (5.6)
No immunosuppressive drug5 (20.8)5 (9.3)NA
Prednisone10 (41.7)28 (51.9)NA
 Median (range) mg/day6.25 (2.5–15)5 (1.25–15)NA
Hydroxychloroquine10 (41.7)30 (55.6)NA
 Median (range) mg/day400 (200–800)400 (200–1,000)NA
Azathioprine6 (25)17 (31.5)NA
 Median (range) mg/day87.8 (50–125)125 (75–200)NA
Methotrexate0 (0)6 (11.1)§NA
SLEDAI, median (range)2 (0–8)2 (0–12)NA
Patient's assessment of disease activity, 0–10-cm VAS, median (range)2.2 (0–5.6)1.6 (0–6.6)NA

Cell-mediated responses against A/H1N1 and A/H3N2 were measured in a subset of vaccinated patients with SLE (n = 38) and control subjects (n = 38) matched for age and sex. This subset was based on availability of a matched control and proper acquisition of PBMCs prior to and 1 month following vaccination. The mean ± SD age of the individuals in this subgroup was 43.4 ± 10.2 years, and 24% were men.

Lower prevaccination cell-mediated responses to A/H1N1 and A/H3N2 in patients with SLE.

In the ELISpot assay, prior to vaccination, patients with SLE had fewer IFNγ spot-forming cells against A/H1N1 and A/H3N2 compared with control subjects (Figure 1). Flow cytometry showed that the frequency of CD4+TNF+ T cells upon A/H1N1 stimulation was lower in patients with SLE than in control subjects (Figure 2B). Patients with SLE also had a lower frequency of IFNγ-positive CD8+ T cells upon A/H1N1 stimulation as well as lower frequencies of IFNγ- and TNF-producing CD8+ T cells upon A/H3N2 stimulation (Figures 3A and B).

Figure 1.

Enzyme-linked immunospot assay of interferon-γ (IFNγ) spot-forming cells per 2 × 105 peripheral blood mononuclear cells in patients with systemic lupus erythematosus (SLE) and healthy controls (HCs) in response to A/H1N1 and A/H3N2 stimulation before vaccination (t = 0 days) and 4 weeks after vaccination (t = 28 days). Results are corrected for responses in unstimulated cultures from the same sample. Bars show the median and interquartile range.

Figure 2.

CD4+ T cell responses against A/H1N1 and A/H3N2. A, Representative example of gating of activated (CD69+) tumor necrosis factor (TNF)–producing CD4+ T cells in a prevaccination sample from a healthy control subject. B and C, Frequencies of cytokine-producing CD4+ T cells upon stimulation with A/H1N1 (B) and A/H3N2 (C) in patients with SLE and healthy control subjects, before vaccination and 4 weeks after vaccination. Results are corrected for responses in unstimulated (Unstim.) cultures from the same sample. Bars show the median and interquartile range. SEB = staphylococcal enterotoxin B; IL-2 = interleukin-2 (see Figure 1 for other definitions).

Figure 3.

CD8+ T cell responses against A/H1N1 and A/H3N2. Frequencies of cytokine-producing CD8+ T cells prior to vaccination upon stimulation with A/H1N1 (A) and A/H3N2 (B) in patients with SLE and healthy control subjects. Results are corrected for responses in unstimulated cultures from the same sample. Bars show the median and interquartile range. For interleukin-2 (IL-2) production following stimulation with A/H3N2 in patients with SLE, both the median and the interquartile range were 0. See Figure 1 for other definitions.

Lower cell-mediated responses to A/H1N1 and A/H3N2 in patients with SLE following influenza vaccination.

Following vaccination, 68.4% of patients with SLE and 71.1% of control subjects showed an increase in the number of IFNγ spot-forming cells against A/H1N1; for A/H3N2, 60.5% of patients and 73.7% of control subjects showed an increase. Increases were similar in patients with SLE and control subjects. After vaccination, the number of IFNγ spot-forming cells remained lower in patients with SLE compared with control subjects (Figure 1).

Following vaccination, the number of A/H1N1-specific IFNγ-producing CD4+ T cells increased in 66.7% of patients with SLE and 65.7% of control subjects. Similarly, the number of A/H1N1-specific TNF-producing CD4+ T cells increased in 61.1% of patients with SLE and 71.4% of control subjects. In 71.4% of control subjects, the number of IL-2–producing CD4+ T cells also increased (Figure 2B). For A/H3N2, 60% of patients with SLE and 61.8% of control subjects showed an increase in IL-2–positive CD4+ T cells following vaccination; 73.5% of control subjects showed an increase in the number of TNF-positive CD4+ T cells as well (Figure 2C). Therefore, in patients with SLE, the response to vaccination was restricted to a more limited cytokine profile. Moreover, patients with SLE reached lower frequencies of TNF- and IL-2–producing CD4+ T cells against A/H1N1 compared with control subjects (P = 0.014 and P = 0.034, respectively).

As was expected, neither patients with SLE nor control subjects showed changes in the percentage of A/H1N1- and A/H3N2-specific CD8+ T cells upon vaccination. Accordingly, postvaccination differences in influenza-specific CD8+ T cells between patients and control subjects were similar to prevaccination differences (data not shown).

Adequate responses of CD4+ and CD8+ T cells following SEB stimulation in patients with SLE.

Upon SEB stimulation, patients with SLE and control subjects showed similar frequencies of IFNγ-, TNF-, and IL-2–producing CD4+ T cells (Figure 4A) and CD8+ T cells (Figure 4B). This indicated that T cells from patients with SLE were generally capable of adequate cytokine responses.

Figure 4.

CD4+ and CD8+ T cell responses to staphylococcal enterotoxin B (SEB). Frequencies of cytokine-producing CD4+ T cell (A) and CD8+ T cells (B) in patients with systemic lupus erythematosus (SLE) and healthy controls (HCs) upon stimulation with SEB in prevaccination samples. Results are corrected for responses in unstimulated cultures from the same sample. Bars show the median and interquartile range. IFNγ = interferon-γ; TNF = tumor necrosis factor; IL-2 = interleukin-2.

Lower antibody response to influenza vaccination in patients with SLE.

Prior to vaccination, patients with SLE had a higher GMT against A/H1N1 as compared with control subjects. One month postvaccination, patients with SLE and control subjects reached comparable GMTs to each vaccine strain. However, the fold increases following vaccination were lower in patients with SLE for the A/H1N1 and A/H3N2 strains. Three to four months after vaccination, titers had decreased in both patients with SLE and control subjects; GMTs remained comparable. Patients with SLE had a lower seroconversion rate for A/H1N1 compared with control subjects (P = 0.001), but for A/H3N2, seroconversion rates in patients with SLE and control subjects were similar. Prior to vaccination, seroprotection rates were comparable in patients with SLE and control subjects. One month after vaccination, patients with SLE had a lower seroprotection rate against the A strains compared with control subjects; this difference was significant for A/H3N2 (P = 0.032). Three to four months after vaccination, seroprotection levels had dropped in patients with SLE as well as control subjects, to comparable levels (Table 2).

Table 2. Hemagglutination inhibition antibodies in patients with SLE and healthy control subjects*
StrainSLE patients (n = 54)Healthy controls (n = 54)
  • *

    SLE = systemic lupus erythematosus; FI = fold increase; seroconversion = ≥4-fold increase in titer; seroprotection = titer ≥40.

  • P < 0.01 versus patients.

  • P < 0.001 versus patients.

  • §

    P < 0.05 versus patients.

Geometric mean titer  
 A/H1N1  
  t = 018.910.9
  t = 28 days (FI)76.5 (4.0)98.2 (9.0)
  t = 3–4 months51.362.7
 A/H3N2  
  t = 015.812.4
  t = 28 days (FI)86.4 (5.5)138.0 (11.1)
  t = 3–4 months55.876.0
Seroconversion rate, no. (%)  
 A/H1N1  
  t = 28 days24 (44.4)42 (77.8)
 AH3N2  
  t = 28 days37 (68.5)41 (75.9)
Seroprotection rate, no. (%)  
 A/H1N1  
  t = 015 (27.8)8 (14.8)
  t = 28 days44 (81.5)48 (88.9)
  t = 3–4 months36 (67.9)39 (72.2)
 A/H3N2  
  t = 08 (14.8)9 (16.7)
  t = 28 days41 (75.9)50 (92.6)§
  t = 3–4 months37 (69.8)45 (83.3)

Taken together, these results indicate that the antibody response in patients with SLE was moderately decreased. This was further substantiated by results in serologically naive patients with SLE and control subjects (prevaccination titer <10). For A/H1N1, 5 (46%) of 11 patients with SLE showed such a seroconversion, versus 20 (80%) of 25 healthy control subjects (P = 0.056); for A/H3N2, this occurred in 1 (14%) of 7 patients with SLE versus 18 (82%) of 22 healthy control subjects (P = 0.003). Finally, we analyzed whether immunosuppressive medication influenced antibody responses. No such influence was observed (data not shown).

Correlations between changes in IFNγ spot-forming cells following vaccination and seroconversions in both patients with SLE and control subjects.

The change in IFNγ spot-forming cells against A/H1N1, measured by ELISpot assay, correlated positively with seroconversion against A/H1N1 (R = 0.311, P = 0.058 for controls; R = 0.348, P = 0.032 for patients with SLE; R = 0.339, P = 0.003 for all vaccinees). For A/H3N2, such a correlation was observed in control subjects (R = 0.318, P = 0.052) but not in patients with SLE. No correlations were observed between CD4+ T cell cytokine responses and antibody responses in control subjects or patients with SLE.

Prior vaccination did not influence cell-mediated responses but did lower antibody responses. In a subanalysis, patients with SLE (n = 13) and control subjects (n = 35) who were not vaccinated in the previous year were evaluated. The groups did not differ in age; the mean ± SD age of patients with SLE was 40.2 ± 8.9 years and that of healthy control subjects was 44.5 ± 9.6 (P = 0.164). In the IFNγ ELISpot assay, patients with SLE had fewer spot-forming cells prior to vaccination against A/H1N1 (P = 0.023) and A/H3N2 (P = 0.034) than control subjects. After vaccination, similar differences were observed, although these differences did not reach significance (P = 0.125 for A/H1N1 and P = 0.051 for A/H3N2). In addition, flow cytometry results showed a tendency toward a restricted CD4+ T cell response in SLE (data not shown).

In this subanalysis, no differences in antibody responses (GMTs, fold increases of GMTs, seroconversion rates, and seroprotection rates) were observed between patients with SLE and healthy control subjects (data not shown). In addition, a comparison was made between patients with SLE who were vaccinated during the previous year (n = 20) and those who were not (n = 34). Vaccination in 2004 led to a higher prevaccination GMT against A/H1N1 compared with those who were not vaccinated in 2004 (26.6 versus 10.5; P = 0.001) and, subsequently, a lowered seroconversion rate (27% versus 75%; P = 0.001).

Treatment with prednisone and/or azathioprine was associated with lower cell-mediated responses to influenza vaccination. Patients treated with prednisone and/or azathioprine (n = 22) were compared with patients who did not receive treatment with these drugs (n = 16). In this subanalysis, no differences were noted prior to vaccination. Following vaccination, patients receiving prednisone/azathioprine had fewer IFNγ spot-forming cells against A/H1N1 and A/H3N2 (P = 0.004 and P = 0.007, respectively) and lower frequencies of A/H1N1-specific IFNγ-, TNF- and IL-2–producing CD4+ T cells (P = 0.004, P = 0.033, and P = 0.036, respectively) as well as A/H3N2-specific IFNγ-producing CD4+ T cells (P = 0.023). No differences in CD8+ T cell responses to A/H1N1 and A/H3N2 were observed (data not shown). In patients not receiving prednisone and/or azathioprine, cell-mediated responses to influenza vaccination were not significantly lower than those in healthy control subjects (data not shown).

No increase in disease activity following influenza vaccination, but more adverse effects in SLE than in control subjects.

Prior to inclusion (Table 1) and during followup, vaccinated and nonvaccinated patient groups did not differ in SLEDAI and VAS scores. At visit 2, the median SLEDAI scores were 2 (range 0–13) in vaccinated patients with SLE versus 2 (range 0–8) in nonvaccinated patients, and at visit 3 the medians were 2 (range 0–10) and 2 (range 0–4), respectively. For VAS scores, the medians at visit 2 were 1.4 (range 0–8.1) in vaccinated patients with SLE and 2.1 (range 0–7.4) in nonvaccinated patients and at visit 3 were 1.8 (range 0–9.4) and 2.2 (range 0–8.9), respectively. Following vaccination, patients with SLE more often reported itching (18% versus 2% of control subjects; P = 0.006), erythema (24% versus 4%; P = 0.003) and induration (30% versus 11%; P = 0.026) at the site of vaccination, and arthralgia (16% versus 4%; P = 0.046). All adverse effects were mild and short-lasting.

DISCUSSION

To our knowledge, this study is the first to evaluate cell-mediated immune responses to subunit influenza vaccine in patients with a systemic autoimmune disease. To perform such an evaluation, we used ELISpot assays and flow cytometry. ELISpot is the more sensitive method, whereas flow cytometry allows phenotyping and detection of multiple cytokines, which offers additional information on the gamma of the response (19).

Cell-mediated recall responses to influenza were lower in patients with SLE. Prior to vaccination, patients with SLE had considerably fewer IFNγ spot-forming cells than control subjects against both A/H1N1 and A/H3N2. CD4+ T cell responses to A/H1N1 were lower in patients with SLE, and this difference reached significance for TNF-producing CD4+ T cells. In addition, CD8+ T cell responses were lower in patients with SLE than in control subjects, for both A/H1N1 (IFNγ production) and A/H3N2 (IFNγ and TNF production).

Following influenza vaccination, cell-mediated responses to influenza remained lower in patients with SLE. Although both patients with SLE and control subjects showed an increase in the number of IFNγ spot-forming cells upon vaccination, for A/H1N1 as well as A/H3N2, the numbers remained lower in patients with SLE. Patients with SLE showed an increase in the number of cytokine-producing A/H1N1-specific and A/H3N2-specific CD4+ T cells following vaccination; however, this increase was restricted with respect to the cytokine profile compared with that in control subjects. Moreover, patients with SLE achieved lower frequencies of A/H1N1-specific TNF-producing and IL-2–producing CD4+ T cells after vaccination. As expected, we did not observe a change in the frequency of cytokine-producing CD8+ T cells following vaccination in either patients with SLE or control subjects.

Because CD4+ and CD8+ T cell responses to SEB were normal in patients with SLE, the decreased cell-mediated response to influenza vaccination could not be attributed to a decreased responsiveness of T lymphocytes in general. Furthermore, the observed differences in cell-mediated responses were, at least largely, independent of previous influenza vaccination status. The rate of influenza vaccination in the previous year was higher in patients with SLE, but in a subanalysis comparing previously nonvaccinated patients with SLE with control subjects, patients with SLE still showed considerably lower responses. Importantly, the use of medications played a major role, because the use of prednisone and/or azathioprine was associated with lower cell-mediated responses against both A/H1N1 and A/H3N2 following vaccination.

A diminished T helper cell response to influenza in patients with SLE, as measured by IL-2 secretion in the supernatant of the influenza-stimulated PBMCs of nonvaccinated patients, has been reported previously (10). We observed a decreased CD8+ T cell recall response to influenza antigens in patients with SLE, which is in accordance with results of a previous study (11). WIV, as used in this study, is able to induce CD8+ T cell responses in vivo and to reactivate memory CD8+ T cells in vitro (ref.20, and de Haan A: unpublished observations). However, WIV might be a weaker stimulus of CD8+ T cells as compared with live virus, due to lower antigen presentation on class I MHC.

Importantly, fewer influenza-specific PBMCs in SLE may be of clinical relevance. Recently, it was shown that the numbers of spot-forming cells correlate with clinical protection from culture-confirmed influenza in young children (21). These numbers may vary depending on the antigen type and influenza strain, because median numbers in our assays were higher than those in assays in which hemagglutinin or vaccine components were used (9, 21–23), and as in the present study, A/H3N2-specific cell-mediated responses were lower than A/H1N1-specific responses. WIV contains core antigens in addition to surface antigens. Also, the uptake and presentation of WIV are more efficient (8). Both factors might contribute to higher responses to WIV compared with hemagglutinin or vaccine components.

Patients with SLE showed normal T cell cytokine responses to SEB. Previous studies demonstrated a normal capacity of PBMCs from patients with SLE to respond to different stimuli, although diminished cell-mediated responses may be present during active disease (10, 24–26). Because our cohort of patients with SLE had predominantly quiescent disease, this may explain their normal responses to SEB. In addition, previous studies showed decreased proliferation of PBMCs (27–29), whereas others showed a normal proliferative capacity (30) or heterogeneous results (31).

Diminished cell-mediated responses to influenza vaccination in patients with SLE appear to reflect, in particular, the effects of immunosuppressive drugs. The effects of previous influenza vaccinations or natural infections could not be completely excluded. Whether intrinsic defects are involved, such as a defective antigen-presenting cell function (32, 33), is uncertain.

In patients with SLE, antibody production upon influenza vaccination is lower than that in the general population (4). In the present study, we too observed lower antibody responses in patients with SLE, as reflected by lower fold increases in titers, a trend toward lower postvaccination GMTs, and fewer seroconversions in serologically naive patients with SLE. Notably, antibody titers are the gold standard for protection, and with regard to seroprotection rates, few differences were observed between patients with SLE and control subjects. Influenza vaccination in the previous year was associated with a lower seroconversion rate to A/H1N1; both vaccines contained the same A/H1N1 strain. The effects of previous influenza vaccination on antibody responses remain subject to discussion, because some studies demonstrated decreased antibody responses (34–36), whereas others showed similar (37–39) or improved responses (40).

We evaluated relationships between antibody and cell-mediated responses, because CD4+ T cell help is necessary for antibody responses (41). However, we did not observe a correlation between CD4+ T cell responses and antibody responses using flow cytometry. We did observe a modest correlation in patients with SLE between changes in IFNγ spot-forming cells against A/H1N1, as measured by ELISpot assay, upon vaccination and seroconversion to A/H1N1. This suggests that in a subset of poorer-responding patients, both cell-mediated and antibody responses are affected. Possibly, no correlation between CD4+ T cell responses and antibody responses was observed due to the lower sensitivity of flow cytometry as compared with ELISpot assay (19). Finally, we showed that influenza vaccination did not induce disease activity over a period of 3–4 months. This confirms the results of previous studies (for review, see ref.5).

Our study has some limitations. First, the sample size was relatively small, and multiple comparisons were made. However, a proper power analysis was not possible, because this study is the first to explore cell-mediated responses to influenza vaccination in patients with SLE. Second, medication use in vaccinated patients with SLE was heterogeneous. Third, more vaccinated patients with SLE than control subjects had received an influenza vaccination in the previous year, which influenced the antibody responses. Fourth, there are no well-defined correlates between cell-mediated responses to influenza and the risk of influenza infection, which limits translation of our results to clinical implications. Fifth, the phenotypes of cells responding in ELISpot assays are unknown. It can be speculated that NK cells are among the cells that have responded in our ELISpot assay (15).

Despite these limitations, we conclude that the combined data point toward diminished cell-mediated immune responses to influenza vaccination in a cohort of patients with SLE representative of those seen in daily practice. Diminished cell-mediated responses may reflect the effects of concomitant use of immunosuppressive drugs. The antibody response to influenza vaccination is also reduced in patients with SLE. Clinicians should be aware that this combined defect might increase the morbidity and mortality due to influenza virus infection, in particular in patients receiving prednisone and/or azathioprine. Therefore, evaluation of clinical protection against influenza in patients with SLE following influenza vaccination seems indicated in order to assess whether more effective influenza vaccines, or vaccination strategies, are warranted.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Holvast had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Holvast, de Haan, Bijl.

Acquisition of data. Holvast, Bijl.

Analysis and interpretation of data. Holvast, van Assen, de Haan, Huckriege, Benne, Westra, Palache, Wilschut, Kallenberg, Bijl.

Acknowledgements

We would like to thank Minke Huitema for optimizing the applied flow cytometry technique.

Ancillary