Julia Singer

Safety and immunogenicity of a novel multivalent OspA vaccine against Lyme borreliosis in healthy adults: a double-blind, randomised, dose-escalation phase 1/2 trial

BACKGROUND Lyme borreliosis is caused by Borrelia burgdorferi sensu stricto in the USA and by several Borrelia species in Europe and Asia, but no human vaccine is available. We investigated the safety and immunogenicity of adjuvanted and non-adjuvanted vaccines containing protective epitopes from Borrelia species outer surface protein A (OspA) serotypes in healthy adults. METHODS Between March 1, 2011, and May 8, 2012, we did a double-blind, randomised, dose-escalation phase 1/2 study at four sites in Austria and Germany. Healthy adults aged 18-70 years who were seronegative for B. burgdorferi sensu lato were eligible for inclusion. Participants were recruited sequentially and randomly assigned to one of six study groups in equal ratios via an electronic data capture system. Participants and investigators were masked to group allocation. Participants received three vaccinations containing 30 μg, 60 μg, or 90 μg OspA antigen with or without an adjuvant, with intervals of 28 days, and a booster 9-12 months after the first immunisation. The coprimary endpoints were the frequency and severity of injection-site and systemic reactions within 7 days of each vaccination, and the antibody responses to OspA serotypes 1-6, as established by ELISA. This study is registered with ClinicalTrials.gov, number NCT01504347. FINDINGS 300 participants were randomly assigned: 151 to adjuvanted vaccines (50 to 30 μg, 51 to 60 μg, and 50 to 90 μg doses), and 149 to non-adjuvanted vaccines (50 to 30 μg, 49 to 60 μg, and 50 to 90 μg doses). Adverse reactions were predominantly mild, and no vaccine-related serious adverse events were reported. The risk of systemic reactions (risk ratio 0·54 [95% CI 0·41-0·70]; p<0·0001) and of moderate or severe systemic reactions (0·35 [0·13-0·92]; p=0·034) was significantly lower for adjuvanted than non-adjuvanted formulations. The 30 μg adjuvanted formulation had the best tolerability profile; only headache (five [10%, 95% CI 4-20] of 50), injection-site pain (16 [32%, 21-45]), and tenderness (17 [34%, 23-47]) affected more than 6% of patients. All doses and formulations induced substantial mean IgG antibody titres against OspA serotypes 1-6 after the first three vaccinations (range 6944-17,321) and booster (19,056-32,824) immunisations. The 30 μg adjuvanted formulation induced the highest antibody titres after the booster: range 26,143 (95% CI 18,906-36,151) to 42,381 (31,288-57,407). INTERPRETATION The novel multivalent OspA vaccine could be an effective intervention for prevention of Lyme borreliosis in Europe and the USA, and possibly worldwide. Larger confirmatory formulation studies will need to be done that include individuals seropositive for Borrelia burgdorferi sensu lato before placebo-controlled phase 3 efficacy studies can begin. FUNDING Baxter.

An exact confidence set for two binomial proportions and exact unconditional confidence intervals for the difference and ratio of proportions

An exact joint confidence set is proposed for two binomial parameters estimated from independent samples. Its construction relies on inverting the minimum volume test, a two-dimensional analogue of Sternes test for a single probability. The algorithm involves computer-intensive exact computation based on binomial probabilities. The proposed confidence set has good coverage properties and it performs much better than the likelihood-based confidence set for the same problem. Applying the principle of intersection-union tests, the method can be used to derive exact tests and confidence intervals for functions of the two binomial parameters. Based on this, new exact unconditional two-sided confidence intervals are proposed for the risk difference and risk ratio. The performance of the new intervals is comparable to that of certain well-known confidence intervals in small samples. Extension of the methods described to two hypergeometric or two Poisson variables is straightforward.

Evaluation of the cellular immune responses induced by a non-adjuvanted inactivated whole virus A/H5N1/VN/1203 pandemic influenza vaccine in humans

In the present study the homologous and heterologous type and subtype specific cellular immune response induced by a wild type inactivated whole virus H5N1 Influenza (A/Vietnam/1203/2004) vaccine was evaluated. Two immunizations with the Vero cell derived H5N1 influenza vaccine on Day 0 and Day 21 induced significant H5N1 vaccine specific and H5 haemagglutinin specific clade and cross-clade reactive CD4(+) T cell responses, which were maintained at significant levels for at least 6 months. The H5N1 vaccine specific response cross-reacted with the H1N1, but not with H3N2 or B seasonal Influenza strains. The vaccine significantly increased the number of H5N1 specific and H5 haemagglutinin specific memory B cells, 6 months after the primary immunization, however no H1N1 specific cross-reactivity was observed. Importantly, the inactivated whole virus H5N1 vaccine was just as effective in inducing CD4(+) T cell and memory B cell response in the elderly (60 years or over) as in the adult population (18-59 years).

Clinical development of a Vero cell culture-derived seasonal influenza vaccine.

BACKGROUND Cell culture technologies have the potential to improve the robustness and flexibility of influenza vaccine supply and to substantially shorten manufacturing timelines. We investigated the safety, immunogenicity and efficacy of a Vero cell culture-derived seasonal influenza vaccine and utilized these studies to establish a serological correlate of vaccine protection. METHODS Two multicenter, randomized, double-blind phase III trials were undertaken in the US during the 2008-2009 Northern hemisphere influenza season, in young (18-49 years) and older (50-64 years and ≥ 65 years) adult subjects. 7250 young adults were randomized 1:1 to receive either Vero-derived vaccine or placebo. 3210 older adult subjects were randomized 8:1 to receive either Vero-derived vaccine or a licensed egg-derived vaccine. Serum hemagglutination inhibition antibody titers were assessed 21 days post-vaccination. Vaccine efficacy in preventing cell culture-confirmed influenza infection was determined for the young adult population. Local and systemic adverse events were recorded in both studies. RESULTS The Vero-derived vaccine was safe and well tolerated in both young and older adults. All US and European immunological licensing thresholds were comfortably met in both populations. Vaccine efficacy in young adults was 79% against A/H1N1 viruses antigenically matching the corresponding vaccine strain and 78.5% for all antigenically matched influenza viruses. A hemagglutination inhibition antibody titer of ≥ 1:15 provided a reliable correlate of protection for the Vero-derived influenza vaccine, with no additional benefit at titers >1:30. Bridging of the correlate of protection established in the young adult population to the older adult immunogenicity data demonstrated the likely effectiveness of the Vero-derived vaccine in the older adult population. CONCLUSIONS A Vero cell culture-derived seasonal influenza vaccine is safe, immunogenic and protects against infection with influenza virus. The novel vaccine technology has the potential to make a substantial contribution to improving influenza vaccine supply. CLINICAL TRIAL REGISTRATION The studies are registered with ClinicalTrials.gov, numbers NCT00566345 and NCT00782431.

A Cell Culture–Derived Influenza Vaccine Provides Consistent Protection Against Infection and Reduces the Duration and Severity of Disease in Infected Individuals

A Vero cell culture–derived seasonal influenza vaccine provides consistently high levels of protection against cell culture–confirmed infection over a complete influenza season. Influenza symptoms are also less severe and of shorter duration in individuals who become infected despite vaccination.

Statistical methods for assessing long-term analyte stability in biological matrices☆

The objective of a long-term stability experiment is to confirm analyte stability in a given biological matrix, encompassing the duration of time from sample collection to sample analysis for a clinical or preclinical study. While long-term analyte stability has been identified as a key component of bioanalytical method validation, current regulatory guidance provides no specific recommendations regarding the design and analysis of such experiments. This paper reviews and evaluates various experimental designs, data analysis methods, and acceptance criteria for the assessment of long-term analyte stability. Statistical equivalence tests based on linear regression techniques are advocated. Both a nested errors and bivariate mixed model regression approach are suitable for application to long-term stability assessment, and control the risk of falsely concluding stability.

A formal comparison of different methods for establishing cut points to distinguish positive and negative samples in immunoassays

Biotechnology derived therapeutics may induce an unwanted immune response leading to the formation of anti-drug antibodies (ADA). As a result the efficacy and safety of the therapeutic protein could be impaired. Neutralizing antibodies may, for example, affect pharmacokinetics of the therapeutic protein or induce autoimmunity. Therefore a drug induced immune response is a major concern and needs to be assessed during drug development. It is therefore crucial to have assays available for the detection and characterization of ADAs. These assays are used to classify samples in positive and negative samples based on a cut point. In this manuscript we investigate the performance of established and newly developed methods to determine a cut point in immunoassays such as ELISA through simulation and analysis of real data. The different methods are found to have different advantages and disadvantages. A robust parametric approach generally resulted in very good results and can be recommended for many situations. The newly introduced method based on mixture models yields similar results to the robust parametric approach but offers some additional flexibility at the expense of higher complexity.

Single priming dose of meningococcal group C conjugate vaccine (NeisVac-C®) in infants

Since the introduction of the meningococcal C conjugate (MCC) vaccine in the pediatric population in 1999, numerous clinical studies have confirmed the immunogenicity and safety of the NeisVac-C(®) vaccine, and several have observed a strong immune response after a single priming dose, which could be successfully boosted. Maximizing protection of infants with as few vaccine doses as possible would increase the general acceptability of the immunization strategies and support broader coverage without increasing vaccination costs. This was a randomized feasibility study of a single priming NeisVac-C(®) vaccine dose administered at 4 or 6 months of age, compared to the currently licensed two dose priming at 2 and 4 months of age, followed by a booster vaccination at 12-13 months of age. High seroprotection rates and serum bactericidal antibody (rSBA) titers were observed in all study groups, whether a single or two dose priming vaccination was administered, at all time points investigated: one month after the priming vaccination(s) (>99% of subjects rSBA≥8), prior to booster vaccination (>65% of subjects with rSBA≥8, with the lowest titers and GMTs seen in the two dose priming group), as well as after booster vaccination administration (99% with rSBA≥128 in all three study groups, with the highest GMT of 2472 seen in the 4 month single dose group). This study confirmed trends seen in previous reports that a single-dose priming vaccination at 4 or 6 months of age can be considered a valuable alternative to the currently licensed two-dose priming vaccination schedule.

Construction of confidence limits about effect measures: a general approach, by G. Y. Zou and A. Donner, Statistics in Medicine 2008; 27:1693-1702.

The authors present a general approach to construct confidence intervals for the difference and ratio of two independent proportions (which can also be extended for two paired proportions). The method described is simple and easy to implement, using only the point estimates and confidence limits of the individual proportions to obtain the confidence limits of the difference or ratio between the two proportions. However, in case of risk ratios the method is presented rather briefly and I think that some further clarification is needed. Also, the conclusion drawn by the authors (‘the method maintains coverage close to nominal’) has to be refined. Table I in the paper (page 1697) presents the performance of the proposed method as a ‘summary of 2000 parameter combinations’. Such a summary misses exactly what is important in these comparisons: the fact that different parameter combinations may behave differently. For instance, the method using Wilson confidence limits has a very low coverage probability for small theoretical proportions (especially in the domain where 1<0.015 or 2<0.015), as it is shown in Figure 1 (the white area corresponds to coverage probabilities exceeding the nominal level). A snapshot from Figure 1 for 1 =0.012 is given in Figure 2. As expected, coverage probabilities improve when the sample size per group increases (Figure 3). I think that the general conclusion ‘the method maintains coverage close to nominal’ needs to be modified into: the method maintains coverage close to nominal when both proportions exceed 1 per cent and the sample size per group is at least 30. When the risk ratio of some rare adverse events is estimated, the poor coverage of the proposed method yields a type I error rate much higher than the nominal one (if inferences are based on the confidence intervals of risk ratios). For example—using the notations from the paper—when (n1, 1)= (20, 0.01) is compared with (n2, 2)= (20, 0.1) at a nominal level of 5 per cent, the null hypothesis of RR= 1 10 is rejected with a true type I error rate greater than 13 per cent. This remains hidden in the summary table presented by the authors. Type I error probabilities for the null hypothesis of RR= 1 10 tested at level 0.05 for different sample sizes are presented in Figure 4. These error rates are close to the nominal level when 2 0.2 for n=20 and when 2 0.1 for n=30.