Immunogenicity of trivalent seasonal influenza vaccine in patients with chronic obstructive pulmonary disease

ABSTRACT

Objective: Current evidence on the immunogenicity of influenza vaccination in patients with chronic obstructive pulmonary disease (COPD) is limited. To address this need for additional knowledge, we conducted a study on the immunogenicity of trivalent seasonal influenza vaccine (TIV) in COPD patients.

Methods: We recruited patients from respiratory outpatient clinics of three hospitals in Tangshan, Hebei province who had stable confirmed COPD, were less than 80 y old, and reported not having had influenza or receiving TIV during the study season prior to enrollment. Patients who had a history of allergy to any TIV component or were classified as having very severe COPD were excluded from the study. Eligible and consenting participants were given one dose of TIV after obtaining a baseline blood sample. A second blood sample was obtained 5 weeks later. We used hemagglutination inhibition (HI) assays to measure antibody responses. We considered seropositive to be an HI titer ≥1:10. We considered seroprotection to be an HI titer ≥1:40 and seroconversion to be either a change from seronegative to a post-vaccination titer of ≥1:40 or a fourfold rise in antibody titer among baseline seropositive subjects. Each subject was followed for 1 month to assess the frequency and type of adverse events.

Results: Eighty-eight subjects completed our study; the median age was 64 y; most (62.5%) had moderately severe COPD; 48.9% of the subjects had comorbid conditions in addition to COPD. Post–vaccination seropositive rates for influenza H1N1, H3N2, and B were all 100%; corresponding seroprotection rates were 96.6%, 93.2%, and 98.9%; seroconversion rates were 81.8%, 87.5%, and 75.0%. There were no statistical differences in seroconversion (P = .10) and seroprotection (P = .30) among the three types of influenza virus. Geometric mean titers (1:) of HI antibodies to H1N1, H3N2, and B were 18.8 (95% CI: 14.0–25.1), 12.2 (95% CI: 9.6–15.4), and 31.8 (95% CI: 26.1–38.8) at baseline, and 267.0 (95% CI: 213.8–333.4), 190.3 (95% CI: 151.7–238.6), and 201.1 (95% CI: 166.5-242.8) after vaccination.

Conclusion: The immunogenicity of one dose of influenza vaccine was excellent in COPD patients. Our study supports recommending influenza vaccination for COPD patients to provide protection from influenza and its complications.

KEYWORDS:

• Immunogenicity
• safety
• trivalent seasonal influenza vaccine
• chronic obstructive pulmonary disease

Introduction

Chronic obstructive pulmonary disease (COPD) is a generally preventable and partially treatable medical condition characterized by persistent airflow limitation. COPD is often progressive and is associated with increased chronic inflammatory responses in the airways and lung tissue to harmful gases or particles such as tobacco smoke.¹ COPD was projected to rank fifth in global economic burden of disease and third in global cause of death by 2020.² Studies conducted in China showed that the prevalence of COPD among individuals 40 y and older was 13.6% in 2014–2015 – a relative increase of 65.9% compared with 2004, that the number of COPD patients increased significantly, and that the disease burden was serious.^3–5 Patients with COPD suffer 0.5 to 3.5 acute exacerbations each year. Exacerbations can lead to a decline in lung function, some of which is irreversible and will accelerate disease processes, increase risk of death, and increase household economic burden.⁶

Seasonal influenza is a common cause of acute exacerbations of COPD.⁶ Influenza vaccine provides partial and variable protection against influenza. Due to genetic drift of influenza viruses and waning immunity following vaccination, seasonal influenza vaccine is updated annually to match anticipated strains and is recommended for annual administration. The vaccine is most effective when given 1 to 2 months before the annual peak of influenza, and is therefore recommended to be given from September through November. When influenza vaccine strains are well matched with circulating strains, the efficacy of the vaccine against laboratory-confirmed influenza disease in healthy subjects under 65 y of age is approximately 70–90%.⁷ Mismatch between vaccine strains and circulating strains is associated with lower effectiveness – approximately 47–77%.⁷ A randomized trial in Thailand has shown that the influenza vaccine is effective at preventing laboratory-confirmed influenza among patients with COPD, with an effectiveness of well-matched strains of 76%.⁸ The World Health Organization (WHO), the US. Centers for Disease Control and Prevention’s Advisory Committee on Immunization Practices (ACIP), and China’s Guidelines for the Diagnosis and Treatment of Chronic Obstructive Pulmonary Disease all recommend influenza vaccination for patients with COPD.

Despite authoritative recommendations, COPD patients are believed to be somewhat immune-deficient compared with healthy people because COPD patients have recurrent bacterial and viral infections⁹ and COPD prevalence increases with age, as does immunosenescence. Due to relative immunodeficiency, COPD patients may have a less effective immune response to vaccination. Current evidence on the immunologic and epidemiologic effects of influenza vaccination for patients with COPD is limited, and most of the relevant studies have been conducted in countries other than China. Studies of influenza vaccine immunogenicity among COPD patients in China are few and have limited quality. We conducted a study of the immunogenicity of trivalent seasonal influenza vaccine in COPD patients in Tangshan, Hebei Province, China to help fill this knowledge need. We report the results of our study and discuss their relevance for protecting COPD patients from seasonal influenza.

Materials and methods

Our study was an intervention cohort study to assess the immunogenicity and safety of trivalent, inactivated, split-virion influenza vaccine.

Setting and subjects

The setting was three hospitals in Tangshan City of Hebei Province: Linxi Hospital, Kangfu Hospital, and Majiagou Hospital. Subjects were recruited from respiratory outpatient clinics of these three hospitals between September and October 2019 using lists of COPD patients from hospital medical record systems.

COPD was graded, and recorded in the medical record, according to patients’ measured FEV₁ as a percent of predicted FEV₁ as mild COPD (FEV₁ ≥ 80% predicted), moderate (50%≤FEV₁ < 80%), severe (30%≤FEV₁ < 50%), and very severe (FEV₁ < 30% predicted).² Doctors from respiratory outpatient clinics invited patients by telephone to participate in the study. If the medical records showed a severity grade assessed within half a year prior to recruitment, that severity grade was used. Otherwise, doctors graded potential subjects prior to recruitment.

To be eligible for the study, patients had to have been diagnosed with COPD and have a ratio of post-bronchodilator, one-second forced expiratory volume (FEV₁) to forced vital capacity (FVC) of less than 0.70;² be less than 80 y of age; have stable COPD; and agree to participate in the study. Patients who had a self-reported influenza illness or received influenza vaccine during September to October 2019, had a history of allergy to any vaccine components, or had a COPD severity grade of “very severe” were excluded.

Vaccination

Participants received trivalent, inactivated, split–virion influenza vaccine, produced by Sanofi Pasteur, that contained vaccine strains recommended by WHO and the European Union for use in the 2019–2020 Northern Hemisphere influenza season. Each 0.5 ml dose contained hemagglutinin (HA) antigens composed of 15 µg of A/Brisbane/02/2018 (H1N1), 15 µg of A/Kansas/14/2017 (H3N2), and 15 µg of B/Maryland/15/2016. The subjects received a single intramuscular injection of the vaccine.

Immunogenicity

We obtained venous blood samples from each subject immediately before and 5 weeks after vaccination. The reason for obtaining the second blood sample 5 weeks post-vaccination rather than the more commonly used 3–4 weeks post-vaccination was because this study was coupled with a pneumococcal vaccination study that required a 5-week sample. A study by Qin and colleagues showed no statistical difference between seropositivity rates and geometric mean titers (GMT) at 1 and 3 months post-vaccination.¹⁰

Sera were separated and stored at −20°C until tested by the National Institute for Food and Drug Control in China (NIFDC). We used hemagglutination inhibition (HI) testing to measure antibody responses in accordance with the WHO protocol for HI testing. Briefly, the serum nonspecific inhibitor was treated overnight at 4°C by a receptor-destroying enzyme derived from Vibrio cholerae, followed by inactivation at 56°C for 50 minutes. Nonspecific agglutinator was removed by adsorption with 50% chicken red blood cells. The test antigen, positive sera, and HA antigen were provided by the National Institute of Biological Products (NIBSC) in England or by WHO.

Consistent with international guidelines for evaluating influenza vaccines,^11,12 seropositivity was defined as an HI antibody titer ≥1:10. Seroprotection was defined as the proportion of subjects with HI titers ≥1:40 5 weeks after vaccination. Seroconversion was defined as either (1) a pre-vaccination HI antibody titer of less than 1:10 and a post-vaccination titer of ≥1:40, or (2) an HI antibody titer ≥1:10 before vaccination and at least a fourfold increase in HI titer after vaccination. For calculating GMT, values below 1:10 (the starting dilution) were assigned as 1:5.

Safety monitoring

Participants were observed 30 min after vaccination for immediate adverse reactions. Subjects were asked to report any adverse events following immunization (AEFI) which was suspected of relating to vaccination, and were contacted by telephone or home visits on days 7 and 28 after vaccination to elicit AEFI reports and answer questions. AEFIs were classified as mild, moderate, severe, or very severe according to the guidelines for adverse event classification standards for vaccine clinical trials, released by the China Food and Drug Administration in 2005 (supplementary materials).

Statistical analysis

For the sample size calculation, we considered the study as a cohort study to determine a point estimate of the proportion of subjects meeting the definition of seroconversion. We assumed that the seroconversion rate in COPD patients would be 70%,¹³ that an acceptable seroconversion rate would be 50%;^11,12 we used a one-sided significance level, α = 0.05, and one-sided power = 0.80 to determine the minimum sample size to be 73. We increased the minimum by 20% to account for the possible attrition, yielding a target sample size of 88.

We used EPI Data3.1 for data entry and SAS9.4 software for statistical analyses. We used chi-square or Fisher’s exact probability testing to compare categorical variables (e.g., seropositive rate, seroprotection rate, and seroconversion rate); we used Student’s t-tests or variance analysis to compare continuous variables (e.g., GMT); P < .05 was considered to be a statistically significant difference.

Ethical review

The study protocol was approved by the institutional ethics committee of the Chinese Center for Disease Control and Prevention (approval number 201940). Subjects provided written informed consent before enrollment.

Results

Subjects

We enrolled 94 subjects; 88 subjects completed the study – 22 from Kangfu Hospital, 43 from Linxi Hospital, and 23 from Majiagou Hospital. Only subjects completing the study were included in the analyses. Table 1 shows characteristics of the subjects, including demographic information and COPD severity. Most were men, with a median age of 64 y, and most had moderately severe COPD; 48.9% subjects had comorbid conditions – mainly hypertension, diabetes, or cardiovascular or cerebrovascular diseases.

Table 1. Characteristics of the study subjects

Variable	Population	Ratio (%)
Age
<65 y	49	55.7
≥65 y	39	44.3
Median (interquartile interval)	64 (59–67)
Range	47–77
Severity of disease
Mild	5	5.7
Moderate	55	62.5
Severe	28	31.8
Gender
Male	65	73.9
Female	23	26.1
Residence
Urban	83	94.3
Rural	5	5.7
Other diseases
Yes	43	48.9
No	45	51.1
Field
Kangfu	22	25.0
Linxi	43	48.9
Majiagou	23	26.1
Total	88	100.0

Immunogenicity

Table 2 shows seroconversion and seroprotection of subjects. Post-vaccination seropositive rates for influenza H1N1, H3N2, and B were all 100%; corresponding seroprotection rates were 96.6%, 93.2%, and 98.9%; seroconversion rates were 81.8%, 87.5%, and 75.0%. There were no statistically significant differences in seroconversion (P = .10) or seroprotection (P = .30) among the three strains of influenza virus. Pre-vaccination seropositivity and seroprotection rates are shown in the supplementary materials.

Table 2. Seroconversion rates and seroprotection rates

		H1N1% (95%CI)		H3N2% (95%CI)		B %(95%CI)
Variable	Cases	Seroconversion	Seroprotection	Seroconversion	Seroprotection	Seroconversion	Seroprotection
Age (y)
<65	49	81.6 (68.0–91.2)	98.0 (89.1–99.95)	85.7 (72.8–94.1)	91.8 (80.4–97.7)	75.5 (61.1–86.7)	98.0 (89.2–99.95)
≥65	39	82.1 (66.5–92.5)	94.9 (82.7-99.4)	89.7 (75.8–97.1)	94.9 (82.7–99.4)	74.4 (57.9–87.0)	100 (91.0-100.0)
Gender
Male	65	81.5 (70.0–90.1)	95.4 (87.1–99.0)	86.2 (75.3–93.5)	90.8 (81.0–96.5)	76.9 (64.8–86.5)	98.5 (91.7–99.96)
Female	23	82.6 (61.2–95.0)	100.0 (85.2–100.0)	91.3 (72.0–98.9)	100.0 (85.2–100.0)	69.6 (47.1–86.8)	100.0 (85.2–100.0)
Severity of disease
Mild and moderate	60	83.3 (71.5–91.7)	95.0 (86.1–99.0)	91.7 (81.6–97.2)	95.0 (86.1–99.0)	78.3 (65.8–87.9)	98.3(91.1–99.96)
Severe	28	78.6 (59.1–91.7)	100 (87.7–100.0)	78.6 (59.1–91.7)	89.3 (71.8–97.7)	67.9 (47.7–84.1)	100(87.7–100.0)
Comorbidities
Yes	43	81.4 (66.6–91.6)	97.7 (87.7–99.9)	81.4 (66.6–91.6)	90.7 (77.9–97.4)	74.4 (58.8–86.5)	97.7 (87.7–99.9)
No	45	82.2 (68.0–92.0)	95.6 (84.9–99.5)	93.3 (81.7–98.6)	95.6 (84.9–99.5)	75.6 (60.5–87.1)	100 (92.1–100.0)
Total	88	81.8 (72.2–89.2)	96.6 (90.4–99.3)	87.5 (78.7–93.6)	93.2 (85.8–97.5)	75.0 (64.6–83.6)	98.9 (93.8–99.97)

Note: There were no statistical differences in seroconversion and seroprotection among the three strains and different groups.

Seroconversion rates for mild- and moderate-severity COPD patients and patients without comorbid conditions were slightly higher than for severe COPD patients and patients with comorbidities, but without reaching statistically significant differences.

Table 3 shows geometric mean HI antibody titers. GMTs (1:) to H1N1, H3N2, and B were 18.8 (95% CI: 14.0–25.1), 12.2 (95% CI: 9.6–15.4), and 31.8 (95% CI: 26.1–38.8) at baseline, and 267.0 (95% CI: 213.8–333.4), 190.3 (95% CI: 151.7–238.6), and 201.1 (95% CI: 166.5–242.8) after vaccination. There were no statistically significant differences in baseline GMTs among the three strains of influenza virus (P = .08). H1N1 GMT was the post-vaccination highest GMT of the strains (P = .03).

Table 3. Geometric mean titers (1:) and 95% confidence intervals

	H1N1		H3N2		B
Variable	Baseline	Post-vaccination	Baseline	Post-vaccination	Baseline	Post–vaccination
Age (y)
<65	21.2 (14.3–31.3)	248.1 (184.2–334.1)	14.9 (10.7–20.7)	184.3 (134.2–253.2)	43.5 (35.4–53.5)^b	244.6 (186.9–320.1)^c
≥65	16.2 (10.3–25.4)	292.8 (207.1–414.0)	9.5 (6.9–13.0)	198.0 (141.7–276.8)	21.5 (15.3–30.1)	157.2 (122.4–201.8)
Gender
Male	18.2 (12.8–25.7)	266.9 (202.3–352.2)	10.9 (8.2–14.5)	183.8 (138.9–243.2)	28.7 (22.4–36.9)^d	200.2 (160.1–250.2)
Female	20.6 (11.6–36.6)	267.1 (185.6–384.3)	16.7 (11.3–24.5)	209.9 (143.4–307.1)	42.5 (32.0–56.5)	203.6 (139.4–297.5)
Severity of disease
Mild and moderate	15.0 (10.9–20.6)^a	245.3 (184.0–327.1)	12.0 (9.2–15.7)	199.3 (150.5–263.8)	30.3 (23.7–38.8)	211.1 (165.6–269.1)
Severe	30.5 (16.6–55.8)	320.0 (227.1–451.0)	12.5 (7.6–20.4)	172.3 (115.2–257.8)	35.3 (24.9–50.2)	181.1 (133.9–244.9)
Comorbidities
Yes	16.0 (10.5–24.3)	247.3 (176.2–347.0)	14.0 (9.9–19.8)	182.0 (127.6–259.7)	26.7 (18.9–37.9)	197.3 (146.7–265.4)
No	21.9 (14.5–33.2)	287.3 (212.6–388.2)	10.6 (7.7–14.7)	198.5 (147.5–267.2)	37.6 (30.7–46.1)	204.7 (159.8–262.3)
Total	18.8 (14.0–25.1)	267.0 (213.8–333.4)	12.2 (9.6–15.4)	190.3 (151.7–238.6)	31.8 (26.1–38.8)	201.1 (166.5–242.8)

Note: a: P = 0.02; b: P = 0.0006; c: P = 0.02; d: P = 0.04. The GMT for influenza virus H1N1 was highest post-vaccination (F = 3.51, P = 0.03).

Compared with subjects ≥65 y of age, younger subjects had higher baseline GMTs (P = .0006) for influenza B and higher post–vaccination GMTs (P = .02). At baseline, H1N1 GMTs among subjects with mild or moderate COPD were higher than among subjects with severe COPD (P = .02); influenza B GMTs were higher among females than males (P = .04). There were no other significant differences in GMT by age, gender, COPD severity, or the presence of comorbidities.

Safety

Two (2.3%) subjects reported mild fever after vaccination. No other AEFIs were reported during the month after vaccination.

Discussion

Our study showed that 5 weeks after administering one dose of trivalent influenza vaccine, seropositivity rates were 100% for influenza H1N1, H3N2, and B; seroprotection rates for the three strains were all above 93%; and seroconversion rates were 81.8%, 87.5%, and 75.0%. Seroprotection and seroconversion were not associated with age, gender, COPD severity, or presence of comorbid conditions. GMTs of antibodies against influenza virus H1N1, H3N2, and after vaccination B were 1:267.0, 1:190.3, and 1:201.1. Reactogenicity of influenza vaccine was very low, with fever being the only adverse event, occurring in less than 3% of the subjects.

In four randomized controlled trials and one observational study that evaluated immunogenicity of trivalent inactivated influenza vaccine in COPD patients,^14–18 seroconversion for H1N1 was 43–80%, seroconversion for H3N2 was 53.1% to 84.1%, and seroconversion for type B was 34.4–61.3%. Seroprotection for H1N1 and H3N2 was ≥76.6% and seroprotection for influenza B was 45.0–72.0%.^14,17,18 The H1N1 and H3N2 seroconversion rates in our study were higher than influenza B seroconversion rates, consistent with other studies.^14,15,18 The seroconversion rates we found were similar to those in Thailand,¹⁴ but higher than seroconversion rates in Australian and United States studies.^15–17 H1N1 antibody titers before and after vaccination were lower than in an Australian study,¹⁷ but GMTs after vaccination were higher than those in Thailand¹⁴ and US studies.¹⁶

We found that immunogenicity was not related to the severity of COPD or the presence of comorbidities, similar to findings in an Australian study.¹⁷ We found no difference in immunogenicity by age, although other studies show that the immune response to influenza vaccine decreases with age.¹⁷

For influenza, immunogenicity is known to be related to protection. There was already evidence showing a protective effect of influenza vaccination in COPD patients. A 2006 Cochrane review that included results of 11 trials found that patients with COPD who were administered inactivated influenza vaccine experienced significantly fewer COPD exacerbations per year than patients with COPD who were given placebo.¹⁹ A retrospective cohort study concluded that the influenza vaccine reduced the incidence of hospitalization from pneumonia and influenza by 52% and resulted in a 70% reduction in all-cause mortality in a cohort of 1898 older individuals with chronic lung disease.²⁰ A Canadian evidence-based review found a 76% reduction of influenza-related respiratory illness due to the administration of trivalent seasonal influenza vaccine.²¹

Possible reasons that some of our findings differed from other studies may be due to variations in background rate of influenza, subject selection criteria, or vaccines studied. The pre-vaccination H1N1 HI antibody titer in our study (1:18.8) was significantly lower than the 1:320 titer in a study conducted in Australia.¹⁷ Our study excluded “very severe” COPD patients and those over 80 y of age, while other studies included these patients.^14,17 The age distribution and range of COPD severity in our study was relatively narrow, as the IQR of age was 59–67 y and most subjects had moderate to severe COPD. We collected blood samples in late December, and some subjects could have had high antibody levels due to natural influenza virus infection.

In healthy adults aged 60 y and over, three studies found seroconversion rates for H1N1, H3N2, and B to be 91.9%, 87.8%, and 81.1%;²² 89.5%, 91.2%, and 87.7%;²³ and >86%, >86%, and 69–72%.²⁴ In our study, 78.4% of the subjects were 60 y or older, and we found slightly lower seroconversion rates. However, the seroconversion rates we found were within the range of acceptability according to the American Biological Products Evaluation and Research Center (the lower limit of the 95% CI for people aged 18–60 should not be less than 40%; the lower limit of the 95% CI interval for people over 60 y old should be not less than 30%).¹¹ The European Committee for Medical Products for Human Use of Seasonal Vaccines specifies that seroconversion rates among 18–60-y-olds should be >40%, and that seroconversion among those over 60 y should be >30%.¹² Thus, the seroconversion rates we found were adequate by these European criteria.

We found very low reactogenicity from vaccination – only 2.3% of the subjects reported any adverse event, and the only adverse event reported was mild fever with a temperature less than 37.6°C. Observed reactogenicity in our study was lower than that in a Thailand study.¹⁴ Possible reasons for lower reactogenicity in our study may be due to different adverse event surveillance methods or to underreporting of adverse events by the subjects in our study. Since we conducted active adverse event monitoring, it is unlikely that any serious adverse events were missed.

The strength of our study was the laboratory testing by NIFDC, which is a WHO Regional Reference Laboratory for influenza. Our study provided data from China for COPD patients, partially filling a knowledge gap. Our study had several limitations. First, we did not have a healthy population for control, so we could not determine whether there is a difference in the immune response between COPD patients and healthy adults. However, because influenza laboratory testing is standardized, the results of published studies on immunogenicity of influenza vaccine in healthy adults allow for a reasonably valid comparison. Second, we did not have a non-vaccinated COPD control group. Influenza vaccination is recommended for COPD patients in China, precluding use of an intentionally non-vaccinated COPD patient control group. Third, we did not select subjects randomly, which may affect representativeness of the subjects. Fourth, our sample size was too small to assess the immunogenicity among subgroups. Fifth, we did not assess influenza illness among subjects during the study. Considering that the normal attack rate of influenza is about 5% of adults per influenza season and that our study period was short, we believe that the possibility of natural infection of subjects was low. Finally, the study was short term and did not provide evidence of persistence of immunity. To our knowledge, only one study showed persistence in COPD patients of influenza vaccine.¹⁴ Future studies could include longer term follow-up to determine the duration of seroprotection.

Future research could refine the relation between immunogenicity and severity of COPD.²⁵ Effectiveness and cost-effectiveness of influenza vaccination in COPD patients following an acute exacerbation could strengthen the evidence base supporting recommendations for influenza vaccine in COPD patients.

In conclusion, the immunogenicity of one dose of influenza vaccine was excellent in COPD patients. Our study supports recommending influenza vaccination for COPD patients to provide protection from influenza and its complications.

Declaration of interests

We declare no competing interests. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Acknowledgments

We thank Jianquan Li, Xiaohui Zhang, Yi Wang, Yanru Zhang, Yadi Su, Xueying Li, and other colleagues for their superb fieldwork. We thank Lance Rodewald, MD, senior advisor to China CDC, for English language editing.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2021.1911515.

Additional information

Funding

This research was funded by the National Key R&D Program of China, grant number 2017YFC1309304; Key Technologies Research and Development Program [2017YFC1309304].