The demographics of the 243 subjects were similar for the LAIV and the TIV recipients in each of the immunologic Groups vaccinated and between the Arms when the three Groups were combined (). CD4% and CD4 count increased progressively from Group 1 to Group 3 in accordance with the experimental design. HIV viral load was similar between the vaccine Arms and between the immunologic Groups.
Demographics of subjects receiving either LAIV or TIV
indicates the adverse events reported within 28 days after administration of each vaccine. The number of subjects with adverse events was similar after either vaccine in all event categories except for the injection site reactions after TIV (23% overall) and more nasopharyngeal symptoms after LAIV compared with TIV (52% vs 31%; p = 0.002). Event frequency did not vary significantly with the immunological Group of the vaccine recipients. Pulmonary signs and symptoms, including the incidence of asthma and wheezing within 28 days after vaccination, were similar in LAIV and TIV recipients.
indicates the occurrence of grade 2 and grade 3 adverse events after each vaccine. No grade 4 events were reported. Grade 2 events occurred in 16–31% of subjects in individual immunological Groups, but there were no statistically significant differences between these Groups, regardless of the vaccine administered. Grade 2 events were more common with TIV as a result of injection site reactions. There were 3 subjects with Grade 3 events following LAIV (one considered vaccine-related) and 2 Grade 3 events following TIV (both considered vaccine-related). The number of ungraded diagnoses were similar in the vaccine Arms. Two cases of pneumonia were reported and radiographically confirmed. One subject who received TIV had a RML pneumonia that occurred 17 days after vaccination and was treated successfully as an outpatient with Azithromycin. Another subject received LAIV 13 days before developing a non-productive cough. Right-sided infiltrates were noted on chest X-ray; white blood cell count was 34,500/µl (74 polymorphs; 5 band forms), and a low-grade fever was present. Nasal wash was negative for influenza by rapid assay and culture, but the PCR demonstrated vaccine strain influenza A (H3N2). The patient improved within 24 hours after an injection of Ceftriaxone.
There were no significant increases from baseline in median/mean plasma HIV viral load after either vaccine in any of the immunological Groups. Similarly, median CD4% did not change significantly at any time point as a result of vaccination.
The immune response to each vaccine was measured by serum HAI assays. indicates the proportion of subjects with HAI titer ≥40 analyzed with the immunologic Groups combined. The proportion with post-vaccination titers of this magnitude against the H1N1 antigen and H3N2 antigen was similar for both LAIV and TIV recipients. TIV recipients were roughly twice as likely to have responses of this magnitude to the B antigen, although the proportion of subjects with this baseline titer was higher in the TIV Arm. A similar proportion of LAIV or TIV recipients also achieved a 4-fold increase in HAI titer against the H1N1 antigen, whereas significantly more TIV recipients achieved a 4-fold increase in HAI titer against the H3N2 antigen at week 4 post-vaccination (but not at week 24) and against the B antigen at both post-vaccination time points. The geometric mean titer (GMT) of specific antibody responses to either vaccine is shown in . The GMT after either vaccine is similar against the H1N1 antigen, but is significantly higher after TIV against the H3N2 and B antigens. Antibody responses to a given antigen in each Arm were similar regardless of the immunological Group (data not shown).
Percentage of Subjects with Titer ≥40 and Percentage with four-fold increase over baseline in HAI Assay
Geometric mean antibody responses to LAIV and TIV
Univariate linear regression analysis of antibody titers in subjects prior to vaccination indicated that HAI GMTs were not significantly different between the three immunological Groups, and that prior to vaccination there was little consistent relationship of HAI GMT to age, CD4% or CD8%, or entry HIV RNA plasma levels. However, the baseline HAI GMT for 3 of 6 baseline comparisons was significantly higher in girls than in boys. Univariate linear regression analysis demonstrated that with all strains in both vaccines the HAI GMT four weeks after vaccination correlated directly with the HAI GMT titer prior to vaccination (P<0.0001). Also at four weeks post-vaccination in LAIV recipients, there was an inverse relationship between entry HIV RNA plasma levels and HAI GMT against one strain (H3N2, P=0.02). In TIV recipients this inverse relationship was observed for all strains (H1N1, P=0.03; H3N2, P=0.05; and B, P=0.004). Age, gender, and number or type of circulating lymphocytes had no consistent effect on HAI GMT after either vaccine.
Multivariate linear regression analyses included the following variables: age, entry log2 (HAI titer), entry CD8%, immunologic Group, and entry HIV RNA. Separate analyses were performed for each strain and each vaccine. These analyses demonstrated direct relationships between the HAI GMT at four weeks after vaccination and entry HAI GMTs for all three influenza strains in both LAIV and TIV vaccines (P<0.0001), and an inverse relationship between entry HIV RNA in plasma and HAI GMT against only two viruses in TIV (H3N2, P=0.04, B, P=0.04). Age, entry CD8%, and immunologic Group were not consistently found in the multivariate regression analyses to correlate with GMT at four weeks after vaccination.
indicates the frequency that vaccine strain influenza was isolated from LAIV recipients. Vaccine strain H1N1 was recovered from 23 of 347 specimens (6.6 %); vaccine strain B was recovered from 11 (3.2 %); and vaccine strain H3N2 was recovered from 3 (<1%). Five specimens contained two vaccine strains. Twenty-five of 30 positive specimens were obtained on days 2 to 4 after LAIV (from 108 specimens). Three of 6 specimens from days 5 and 6 contained vaccine virus. The mean titer of shed virus for both H3N2 and B strains was approximately 2.0 log10 TCID50, and was 2.6 log10 TCID50 for H1N1. Two of 105 specimens taken on days 11 to 15 were positive for type B in a primary isolation at the U. of Colorado lab, but could not be identified as vaccine strain or wild type viruses due to insufficient virus in the sample. Neither of these isolates came from a subject that had virus isolated at an earlier time. None of 128 additional specimens obtained after day 15 were positive for vaccine virus. One wild-type virus was isolated on day 14 and another on day 28; both were H3N2 (data not shown).
Shedding of the H1N1 LAIV at day 3 was significantly associated with low baseline HAI antibody against that virus (p<0.01), and there was also a trend for this association with shedding of the H3N2 virus (p = 0.07). Shedding, examining all viruses combined or each individually, did not correlate with age, CD4 count or CD4%, or HIV viral load at the time of vaccination, and did not correlate with the subsequent boost in any specific antibody measured at 4 weeks post-vaccination.