EU Regulatory Perspective on RSV vaccines

Transcription

1 EU Regulatory Perspective on RSV vaccines

2 EU Perspective No RSV-specific guideline in the EU There is no established EU (i.e. CHMP) position CHMP scientific advice has been given, which reflects current EU regulatory thinking in the absence of specific guidance Expected to file EU MAAs All the efficacy data may come from LMICs Expected to be centralised MAAs Products of genetic engineering (compulsory) Usual route for all new vaccines

3 Topics Topics considered Maternal immunisation and determination of efficacy in offspring (± primary vaccination of infants) Primary vaccination in infants Safety database and duration of safety follow-up Topics not considered Primary vaccination in all other age groups, whether healthy or at risk Details of assessing immune responses

4 Practical issues for Sponsors and Regulators Assays to detect RSV breakthrough cases Case definitions; virological plus clinical criteria Case ascertainment; implications of seasonality Maternal vaccination /passive protection of infants Infant vaccination Age de-escalation issues Duration of protection Safety database and duration of follow-up

5 Assays for detection of RSV Culture depends on controlling for viral thermolability and results depend on laboratory experience Antigen detection sensitivity may be 80-90% vs. culture Prefer RT-PCR assays; avoid in-house assays if possible Select a single commercial RT-PCR assay (may not be approved for routine use in all participating countries) Need for and feasibility of central laboratory confirmation depends on the assay selected and the study sites Specimen collection, storage and shipping issues

6 Assays for detection of RSV some questions Do we want to know if it is RSV A or B? Do we want to know if there is co-infection (e.g. with human metapneumovirus), which may be associated with more severe disease and so could confound the primary endpoint if the aim is to prevent severe RSV disease? Is there a need to look at virus from breakthrough cases (e.g. to look for escape mutants) by genotyping? Do we want an assay that can quantify viral load in samples? Do we want to document duration of shedding in breakthrough cases to assess any shortening by vaccination?

7 Case definitions - what will the vaccine prevent? Is the aim to prevent any RSV-associated clinical illness or to prevent specific presentations? For example, cases that: Meet any or selected/defined respiratory symptoms Involve acute lower respiratory illness (ALRI) necessitating contact with a healthcare professional (HCP) Require a GP visit / clinic visit / ER visit / hospitalisation Require supplementary O2 based on saturation or PaO2 Require ventilation (any or PPV) Occur in healthy children and/or in at-risk children

8 Case ascertainment Need active ascertainment to check that all cases counted Exact method depends on the case definition Several concerns regarding: o What will be trigger for HCP contact (need to define) o What will be the route to HCP (parental decision only; contact with site staff first; contact with GP first; access to HCP in and out of hours; distance to HCP; bypass cases [direct presentation to hospital]; decisions to admit may be very variable across sites due to the set-up of the healthcare services Although there will be a placebo control there could be lack of homogeneity of cases counted in primary analysis

9 Vaccinating in pregnancy Clinical trial approval based on safety for mother and fetus EC and/or PH authorities may require/desire data to answer the question regarding any direct benefit to mothers Trials will not be powered to estimate VE in vaccinees Little information on rate of clinically apparent RSV infections and severity in pregnant women Unlikely that numbers of cases of RSV-associated clinical illness in mothers during last trimester ± post-partum will provide definitive estimate of direct benefit

10 When to vaccinate pregnant women in trials Maternal vaccination will be in 3 rd trimester (> week 26) Vaccination in the last 3 months of the RSV season means the infant is unlikely to be exposed to RSV before maternal antibody declines below protective levels Issues for clinical trial design, initiation and recruitment To maximise the likelihood of accumulating enough cases in the control group, could target women expected to give birth during the RSV season when minimum # months left The RSV season (or lack of seasonality) may not be welldefined at potential study sites; may need a pre-study

11 Passive protection of infants Many factors may influence the maternal immune response AND the efficient transfer of IgG across the placenta Need to decide if the first efficacy studies will be broadly inclusive (in which case likely need to stratify) or will exclude sub-groups most likely to have poor trans-placental transfer Need to assess transfer by measuring cord blood antibody Cord blood can be obtained only if the mothers give birth in settings in which samples can be obtained and then handled, stored and shipped in an appropriate fashion Interactions of these factors will influence site locations

12 Passive protection of infants Cases counted in the primary analysis may be confined to infants born a minimum number of weeks after maternal vaccination and of a minimum gestational age at birth Use of palivizumab will confound the endpoint - need to understand the actual mode of palivizumab use across the study sites to define cases for the primary analysis Variable national use of palivizumab despite the official SmPC [EU for infants born at 35 weeks gestation and < 6 months of age at the onset of the RSV season or aged < 2 years and treated for bronchopulmonary dysplasia within last 6 months or with haemodynamically significant congenital heart disease]

13 Passive protection of infants Protection may reflect not only titre at birth but the rate of decline, which may not be constant in all settings Assumption then follows that the median time to breakthrough infections will differ in subsets In a case-driven study the primary analysis could reflect a large contribution of cases from a country/region with poor transfer and rapid decay of maternal antibody The statistical analysis plan will need to deal with any predictable covariates and include plans to explore covariates

14 Vaccinating infants - Age de-escalation Initial studies can be done in adults but they will not be RSVnaïve; limited or negligible value for prediction of safety and immunogenicity in infants RSV becomes a mild disease as age increases so benefit of vaccination may be minimal beyond the first few years of life A typical age de-escalation approach is likely not feasible in healthy children; perhaps could start with a mini-cohort in the upper part of the target age range In at-risk children it may be possible to start at a higher age but probably non-naïve so not so relevant to healthy infants

15 Vaccinating infants Aim for strong neutralising antibody response associated with a non-th 2 dominant immune response Need to assess this in RSV-naïve infants (need cells) Such findings cannot rule out risk of disease enhancement Case definition and mode of ascertainment could be the same as for passive protection; same issues apply Additional endpoints of interest could be duration of RSV shedding and/or effects on viral load in breakthrough cases in case vaccine attenuates (e.g. using quantitative PCR)

16 When to vaccinate infants? In infants born to vaccinated mothers Need to estimate the duration of passive protection before deciding when to start active vaccination in infants based on efficacy and not antibody persistence Practicalities of infant vaccination against RSV after completion of the routine primary series (e.g. extra visits) Should primary series start as early as seems needed? Should primary series start also take into account when the RSV season starts/ends if there is seasonality?

17 When to vaccinate infants? In the absence of maternal vaccination Start primary series as early as possible in infancy? In both scenarios Essential to investigate whether/to what extent maternal antibody reduces the response to active vaccination in infants An important inhibitory effect of maternal antibody: Has implications for extrapolations between populations May result in different minimum age for regions or subsets (inevitably an approximation since it will be totally impractical to link to individual cord blood titres)

18 Vaccinating infants - Duration of protection RSV becomes a mild disease as age increases unless the child has underlying risk factors Vaccinated infants should be followed through at least one season for the primary analysis Longer-term follow-up to rule out disease enhancement and consideration of need for boosters in healthy children; the case definition may need to change with age; how long? The investigation of protective efficacy beyond a certain age could be confined to the at-risk group who are kept under medical review and may benefit from booster doses?

19 Safety database for a novel vaccine The current EU general expectation is a minimum of 3000 exposed to the final dose regimen of the vaccine This may be acceptable for maternal immunisation subject to lack of any biological plausibility of adverse effect on fetus The safety database for infants will require further thought Likely to be construct-specific Need enough evidence (from a combination of nonclinical and clinical investigations) to support negligible risk of disease enhancement after vaccination of the RSV-naive

20 Summary The sponsor should decide the primary objective of the programme, i.e. decide on the target indication The sponsor should define the primary case definition and mode of case ascertainment accordingly Pivotal studies must show superiority for vaccine vs. no vaccine against RSV over at least one season; the magnitude of vaccine efficacy expected depends on the population The timing of vaccination needs careful consideration Lots of issues may affect where sponsors choose to do their studies and the validity of extrapolating the findings