Immunogenicity and safety of different dosing schedules of trivalent inactivated influenza vaccine in pregnant women with HIV: a randomised controlled trial

Background: Standard-dose, seasonal, trivalent, inactivated influenza vaccine induces moderate-to-low haemagglutination-inhibition antibody responses in people living with HIV. This study assessed the immunogenicity and safety of different dosing schedules of inactivated influenza vaccine in pregnant women living with HIV in South Africa. Methods: In this double-blind, randomised, controlled trial, we recruited pregnant women with HIV from seven antenatal clinics in Soweto, South Africa. Pregnant women were eligible if they were aged 18–38 years, infected with HIV, and had an estimated gestational age of 12–36 weeks. Women were randomly assigned (1:1:1), using a computer-generated randomisation list, to receive inactivated influenza vaccine containing 15 μg of each of the three seasonal influenza strains for that year, as a single dose, a double dose, or two single doses 1 month apart. Participants and study personnel were masked to group allocation. Haemagglutination-inhibition antibody responses were measured for all groups in the mothers at enrolment and at 1 month after each vaccine dose, and in the single-dose and double-dose groups within 7 days of birth in the neonates. Immunogenicity analyses only included women with visits 28–35 days apart and infants who were born at least 28 days after maternal immunisation. The primary was seroconversion rate to each of the vaccine strains in the mothers 1 month after completion of the dosing schedule, and the primary safety outcomes were frequency of local and systemic reactions. Safety was assessed in mothers and infants until 24 weeks post partum and analysed in all participants who received at least one dose of vaccine. This study is registered with ClinicalTrials.gov, NCT01527825, and is closed to accrual. Findings: Between Feb 11, and June 6, 2013, 800 pregnant women living with HIV were enrolled and randomly assigned to the single-dose (n=266), double-dose (n=265), or two-single-doses (n=269) group. In the analysable population, seroconversion rates in mothers 1 month after the final vaccine dose were significantly higher in the double-dose group (n=230; ranging from 29% to 65% for the three vaccine strains) than in the single-dose group (n=230; ranging from 18% to 49%; p≤0·019 for the three vaccine strains), but were similar between the two-single-doses group (n=220; ranging from 23% to 52%) and the single-dose group (p≥0·20 for the three vaccine strains). Safety outcomes were similar in the three groups, except for more injection-site reactions in recipients in the double-dose group. Interpretation: A regimen of double-dose inactivated influenza vaccine gave slightly greater immunogenicity than did a single-dose regimen in pregnant women living with HIV. However, immunogenicity in the double-dose group was still lower than historical data from the same setting in pregnant women without HIV. More immunogenic vaccines are needed for pregnant women living with HIV to enhance transplacental transfer of vaccine-induced protective antibodies to their newborn infants. Funding: Bill & Melinda Gates Foundation. In this double-blind, randomised, controlled trial, study staff screened pregnant women with documented HIV-1 infection for study participation at seven antenatal clinics in Soweto, South Africa. Soweto is an urban low-income setting with a predominantly black African population of an estimated 1·2 million people located in the outskirts of Johannesburg. HIV prevalence among pregnant women in Soweto has remained stable at 29% since 2009.19 Pregnant women were eligible if they were aged 18–38 years, had an estimated gestational age of 12–36 weeks, and were able to understand and comply with the study protocol. Exclusion criteria included any features of WHO clinical category 3 or 4 of AIDS, having been given any other inactivated influenza vaccine during the current influenza season or received any live licensed vaccine in the past 28 days or an inactivated licensed vaccine in the past 14 days before the study. Full exclusion criteria are in the appendix (p 2). We enrolled participants before the anticipated onset of the influenza season in South Africa. The 2013 South African influenza epidemic period was from April 22 to Oct 13.20 The study was approved by the Human Research Ethics Committee of the University of the Witwatersrand (111114) and done in accordance with Good Clinical Practice guidelines. Mothers provided written informed consent for themselves and their infants. The study protocol is available in the appendix (pp 10–70). Participants were randomly assigned (1:1:1) to receive treatment with a single dose of inactivated influenza vaccine followed by placebo (single-dose group), a double dose of inactivated influenza vaccine followed by placebo (double-dose group), or two single doses of inactivated influenza vaccine (two-single-doses group). The second injection for each group was given 1 month after the first injection. An unmasked study statistician created computer-generated randomisation lists in blocks of 30, with ten blocks in each group (single-dose group, double-dose group, and two-single-doses group), with assignment of consecutive four-digit study numbers to the randomisation list. The randomisation forms were preprinted with the four-digit study numbers and an alphabetical and colour code for vaccine and placebo. Participants were allocated consecutive study numbers in order of enrolment at each study clinic. Participants and study personnel were masked to group allocation. Study medication was prepared by an unmasked pharmacist in syringes filled with either saline or inactivated influenza vaccine as per randomisation and labelled as dominant and non-dominant, and study nurses who enrolled the participants selected syringes matching the codes reflected in the randomisation form (appendix p 3). The pharmacist and study statistician were not involved in participant follow up. Participants were given a commercially produced influenza vaccine (Vaxigrip; Sanofi-Pasteur, Lyon, France) that contained 15 μg of A/California/7/2009 (A/H1N1pdm09), A/Victoria/361/2011 (A/H3N2), and B/Wisconsin/1/2010 (Yamagata-lineage), as recommended for the southern hemisphere in 2013. Vaxigrip was approved in South Africa for use as a single dose in adults. The study pharmacist prepared inactivated influenza vaccine or sterile 0·9% normal saline solution (placebo) in identical syringes; the two preparations were visually indistinguishable. At enrolment (visit 1), all participants received two injections: inactivated influenza vaccine and placebo were given to the single-dose and two-single-doses groups, one in each arm, with inactivated influenza vaccine given in the non-dominant arm; and one dose of inactivated influenza vaccine was given in each arm of the participants in the double-dose group (ie, 30 μg of antigen per strain). 1 month later (visit 2), all participants received a single injection on their non-dominant arm, with the single-dose and double-dose groups both given placebo, and the two-single-doses group given another dose of inactivated influenza vaccine. Participant follow-up was done at the Respiratory and Meningeal Pathogens Research Unit, Soweto, South Africa. We tested immune responses on plasma samples collected from the women immediately before the first injection at visit 1, 1 month after the first injection and immediately before the second injection at visit 2, and 1 month after the second injection at visit 3. We also collected blood within 7 days of birth from the women and their neonates. After review of maternal haemagglutination-inhibition antibody titres, and that no differences were detected between the single-dose and two-single-doses groups, we made a post-hoc decision to analyse only infant blood collected from infants in the single-dose and double-dose groups. Haemagglutination-inhibition antibodies were measured as described previously.5, 21 We did weekly active surveillance by home visit or telephone call from the time of enrolment through to 24 weeks post partum in mothers and infants for influenza-like illness. The criteria for diagnosing influenza-like illness in the women were onset in the past 7 days of symptoms including: a fever of 38·0°C or higher on oral measurement or history of chills, rigors, or feeling feverish; presence of cough, sore throat, or pharyngitis; presence of myalgia, arthralgia, or headache; or presence of dyspnoea, breathing difficulty, or chest pain when breathing. Criteria used for diagnosing influenza-like illness in infants were any of the following: axillary temperature of 37·8°C or higher or mother’s perception that an infant was feverish, or both, without evidence of a non-respiratory localised source, coupled with at least one sign or symptom of acute respiratory illness in the past 72 h; or at least two signs or symptoms of acute respiratory illness in the past 72 h. Signs or symptoms of acute respiratory illness were respiratory rate of 60 or more breaths per min in infants aged 0–2 months and 50 or more breaths per min in infants aged 2–6 months and difficulty breathing reported by the mother, cough, wheezing, runny or congested nose, cyanosis or oxygen saturation below 90% on room air, chest wall in-drawing, grunting on expiration, and pus draining from either ear. Women and infants identified as having influenza-like illness during surveillance visits had respiratory secretions collected within 72 h of illness identification that were tested by influenza PCR. We also investigated by influenza PCR, participants who were admitted to hospital and who had unsolicited study clinic visits due to respiratory illness. Details on sample collection, influenza diagnostic and subtyping by PCR have been described previously.5 For routine assessment of reactogenicity and safety, the investigator (usually study nurse) observed each participant for 30 min after each injection to monitor immediate adverse events. Additionally, participants were provided with a diary card to document solicited local and systemic reactions for 7 days after each vaccination visit. For injection site reactions, severe reactions were defined as follows: severe tenderness as severe pain in the injected limb that increases when moved or when the movement was reduced, severe redness as redness in an area of 100 mm or larger in size, severe swelling as swelling in an area of 100 mm or larger in size, severe hardness as a large lump being felt wider than half the arm width, severe bruising as a bruise larger than 25 mm in size, and severe itching as itching that requires soothing cream. For systemic reactions, severe reactions were defined as severe weakness or tiredness as being unable to do normal activities during the day, severe headache as a headache requiring medication and being unable to do normal activities during the day; severe fever as an axillary temperature of 39·4°C or higher, and severe joint or muscle pain as severe aching that required medication and that restricted activity. Serious adverse events were collected throughout the study until 24 weeks post partum in women and infants. The coprimary study objectives were to assess the relative immunogenicity of a double dose and two single doses of inactivated influenza vaccine compared with a single dose of inactivated influenza vaccine in pregnant women living with HIV to each of the three vaccine strains, and to assess the relative safety of the three dosing schedules. The primary outcome was the seroconversion rate to each of the vaccine strains in mothers 1 month after completion of the dosing schedule. Secondary objectives reported here were to compare the proportion of neonates born to women in the double-dose group with haemagglutination-inhibition titres of 1/40 or higher at age 7 days or younger (individually for each of the three vaccine strains) with the proportion in the single-dose group, to assess the relative efficacy of the double-dose and two-single-doses regimens compared with the single-dose regimen against PCR-confirmed influenza among the women and their infants until 24 weeks post partum. Additional secondary objectives (appendix p 3) will be reported in the second half of 2020. Criteria for interpreting haemagglutination-inhibition results included titres of 1/40 or higher as putative measures of relative seroprotection and seroconversion if a four-fold or higher titre increase occurred from baseline to after vaccination with haemagglutination-inhibition antibody titres of 1/40 or higher after vaccination. Primary safety outcomes were the frequency of solicited local and systemic reactions after a single vaccine dose, a double vaccine dose, or two single doses. We powered the sample size at 80% with a two-tailed significance level of 0·05 to detect a 30% or higher difference in the proportion of pregnant women living with HIV who seroconverted in the double-dose group or the two-single-doses group, compared with the single-dose group. On the basis of results from our previous study on immune responses to a single dose of inactivated influenza vaccine5, we estimated that approximately 44% of women in the single-dose group would have haemagglutination-inhibition antibody titres of 1/40 or higher and a four-fold or higher increase in titres to the least immunogenic strain in the vaccine. The sample size required was 789 women (ie, 263 in each group). We only included women in the immunogenicity analysis if their vaccination visits were 28–35 days apart and if blood samples were collected 28–35 days after the second injection for those in the two-single-doses group. We only included infants in the immunogenicity analysis if blood samples were collected in the first 7 days of life and if they were born at least 28 days after maternal immunisation. We compared proportions with χ2 or Fisher’s exact tests and demographic continuous variables with Student’s t test or the Mann-Whitney test. We estimated geometric mean titres and corresponding 95% CIs by use of logarithmic transformation and compared these between study groups using Student’s t test. We calculated relative vaccine efficacy against first episode of PCR-confirmed influenza in the intention-to-treat population with the formula (1–Isg)/ISD, where Isg is the incidence of cases in the double-dose or two-single-doses group, and ISD is the incidence of cases in the single-dose group, and we calculated 95% CIs and we tested between-group differences. The safety analysis set comprised participants who received at least one dose of vaccine. After the first vaccination, we compared the reactions of women who had been given a single dose of vaccine (the single-dose and two-single-doses groups) with those who had been given a double dose (double-dose group), and after the second vaccination we compared women who had received placebo in the single-dose group only with those who had been given another dose of vaccine (the two-single-doses group). We present safety data as incidences with 95% CIs. We did descriptive analyses for the number of maternal, infant, and fetal deaths, and maternal and infant admissions to hospital. We considered p values of less than 0·05 to be significant. We collected and managed study data using Research Electronic Data Capture (REDCap).22 We did all analyses using STATA version 12.1 (College Station, TX, USA). The study was registered at ClinicalTrial.gov, {“type”:”clinical-trial”,”attrs”:{“text”:”NCT01527825″,”term_id”:”NCT01527825″}}NCT01527825. The funder of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.