Should Pneumococcal Serotype 3 Be Included in Serotype-Specific Immunoassays?
Round 1
Reviewer 1 Report
Of all pneumococcal serotypes, serotype 3 maybe is the odd one out. In early stduies in mice, serotype 3 was used as prototype serotype for immunological mechanisms and vaccine development. For the human immune system, serotype 3 always has been an outlier.
This manuscript give an overview of serotype 3 conjugate vaccine efficacy (low to absent), speculates about potential mechanisms and mentions the difficulty is measurment of antibodies. In my view, latter problems are not (directly) related to vaccine inefficcay. The suggestion to stop with measuring the antibodies therefore is not well argumented. For correct diagnosis of a current patient with a serotype 3 IPD, it is important to know whether such a patient is able to produce a specific antibody response. If technical problems arise, those should be tackled.
Specific comments:
line 151: to me it is not clear why a thick capsule should interfere with the ability of antibodies to bind and opsonize.
Figure 1: please explain what is on the X-axis (individual serum samples, individual particpating lab?)
The references is a mixture of formats: authors either have a full first name, an initial, or nothing
Numbering of headings: 2 times #4.
Author Response
This manuscript give an overview of serotype 3 conjugate vaccine efficacy (low to absent), speculates about potential mechanisms and mentions the difficulty is measurment of antibodies. In my view, latter problems are not (directly) related to vaccine inefficcay. The suggestion to stop with measuring the antibodies therefore is not well argumented. For correct diagnosis of a current patient with a serotype 3 IPD, it is important to know whether such a patient is able to produce a specific antibody response. If technical problems arise, those should be tackled.
The central thesis of this paper is that the effectiveness of conjugate vaccines against serotype 3 is limited, evidence to support this statement is provided, along with evidence that the currently accepted correlate of protection is far too low. It is the opinion of the authors that, technical difficulties with the assay aside, the result for serotype 3 is of little to no clinical relevance as it is clearly established that production of antibody above the currently accepted correlate of protection does not necessarily confer immunity, and that negative results provide no information on the patient’s intrinsic ability to mount an antibody response given the poorly understood response to the vaccine.
We have expanded the discussion to make clearer the fact that evidence from the UK NEQAS showing poor reproducibility is somewhat incidental to the central thesis of the paper – the measurement itself is of no clinical value, and thus the effort required to address the evident technical issues is not worth pursuing in depth.
Specific comments:
line 151: to me it is not clear why a thick capsule should interfere with the ability of antibodies to bind and opsonize.
Interestingly, several studies exist demonstrating this phenomenon (we have added an additional citation to support this). A similar study in Spain has also seen the same phenomenon with meningococcal serogroup C with increased production of capsule (Uria, Maria Jose et al. “A generic mechanism in Neisseria meningitidis for enhanced resistance against bactericidal antibodies” Journal of experimental medicine vol. 205,6 (2008): 1423-34.)
Figure 1: please explain what is on the X-axis (individual serum samples, individual particpating lab?)
We have added a label to the x axis showing individual samples from each NEQAS distribution. N.B. As the legend states, the white bars show antibody GMC and the black bars show the %agreement between the 5-9 participating laboratories.
The references is a mixture of formats: authors either have a full first name, an initial, or nothing
We have fixed this. The authors naively assumed that automated bibliography software might have improved recently.
Numbering of headings: 2 times #4.
We have fixed this also. Many thanks for pointing this out.
Reviewer 2 Report
The commentary by Linley et al. discusses the efficacy of current pneumococcal vaccines in protecting against serotype 3 invasive disease, carriage and the variability in results of the Luminex assay.
1. The title focuses on the antibody assay while the assay is only one small part of a communication on pneumococcal serotype 3. Perhaps the title should be changed?
2. Invasive pneumococcal disease: While PCV13 and other pneumococcal vaccines have been quite unsuccessful in causing reduction of invasive serotype 3 disease comparable to other vaccine serotypes, not all studies have shown no efficacy as suggested in the review. A review by De Wals in Vaccine 2018, volume 36, p 5495-5496 describes the range in efficacy and also the limitations in study designs and inter-study comparisons relating to vaccine type, age group, vaccine schedule, etc. While this is briefly touched upon in the present manuscript the section on IPD could be better balanced to give the reader a better understanding of differences between studies.
3. The authors do not mention vaccine efficacy against otitis. Since they discuss both IPD and carriage this seems like a gap in the manuscript, especially since AOM is discussed in the first paragraph of section 4. Lewnard et al., for example, discuss the experience in both carriage and otitis media in Clinical Infectious Diseases 2017;65; 1853-61.
4. Luminex assay reproducibility (should perhaps be “5.” As there is a preceding section 4?). This section needs much clarification. A) If the results have been previously published they need to be properly referenced; B) If not previously published, the methods, inter institution comparison, standard deviations, statistical analyses etc. need to be presented and the associated study limitations discussed. Also, there is no comparison to assays or results from other countries.
5. The conclusion could be more precise. A) The suggestion that the serotype 3 component part of the immune assay should be removed is not supported by the data in the present form as the Luminex section is too brief. B) Do the authors suggest the serotype 3 component is useless and should therefore be removed from vaccines or that the efficacy needs to be better analyzed and that better understanding of serotype 3 could lead to a better incorporation of this serotype in a future vaccine formulation?
Author Response
1. The title focuses on the antibody assay while the assay is only one small part of a communication on pneumococcal serotype 3. Perhaps the title should be changed?
This paper was prompted by discussions amongst those using the Luminex assay to provide a clinical Pneumococcal serotype-specific antibody measurement service. The aim of the paper is to summarise evidence regarding the limited clinical usefulness of serotype 3 measurements. We do not believe that a change of title is appropriate.
2. Invasive pneumococcal disease: While PCV13 and other pneumococcal vaccines have been quite unsuccessful in causing reduction of invasive serotype 3 disease comparable to other vaccine serotypes, not all studies have shown no efficacy as suggested in the review. A review by De Wals in Vaccine 2018, volume 36, p 5495-5496 describes the range in efficacy and also the limitations in study designs and inter-study comparisons relating to vaccine type, age group, vaccine schedule, etc. While this is briefly touched upon in the present manuscript the section on IPD could be better balanced to give the reader a better understanding of differences between studies.
We have added in the two studies showing positive effectiveness of PCV13 against serotype 3, along with the further Spanish study showing no effectiveness, and have discussed the paper by De Wals. We feel that De Wals conclusions, that contradictory evidence of these studies demonstrate rapidly waning antibody response to serotype 3 vaccination, serves to strengthen the evidence that measurement of specific serotype 3 antibodies is of very limited clinical relevance.
3. The authors do not mention vaccine efficacy against otitis. Since they discuss both IPD and carriage this seems like a gap in the manuscript, especially since AOM is discussed in the first paragraph of section 4. Lewnard et al., for example, discuss the experience in both carriage and otitis media in Clinical Infectious Diseases 2017;65; 1853-61.
We appreciate the reference provided and have now edited the text to include it. However, this paper does not include per serogroup AOM incidence or carriage data (nor do those studies cited within), only the calculated rates of progression, and so discussion of this paper could not be extended to the carriage section. We have moved discussion of Cohen et al.’s work on carriage in AOM patients to this section. The authors note that studies examining the serotype specific efficacy of PCV13 against OM are limited.
4. Luminex assay reproducibility (should perhaps be “5.” As there is a preceding section 4?). This section needs much clarification. A) If the results have been previously published they need to be properly referenced; B) If not previously published, the methods, inter institution comparison, standard deviations, statistical analyses etc. need to be presented and the associated study limitations discussed. Also, there is no comparison to assays or results from other countries.
As stated in the text, these data are taken directly from the UK NEQAS IIA programme with their permission. These data are not “published”, they taken from NEQAS Distributions, which are not made available to non-participants, and so citation of these data was not thought appropriate. Information regarding the analyses used is freely available on the UK NEQAS IIA website, we feel it is beyond the scope of this short communication to include a full description of the UK NEQAS IIA methods, these will be familiar to the intended audience. A reference to the UK NEQAS IIA website has been added.
Participants in this programme are from both the UK and overseas- it is an international programme. Information regarding the identities of the participating laboratories is not released by NEQAS, so this cannot be included. The fact that the programme is international has been added to the text.
5. The conclusion could be more precise. A) The suggestion that the serotype 3 component part of the immune assay should be removed is not supported by the data in the present form as the Luminex section is too brief. B) Do the authors suggest the serotype 3 component is useless and should therefore be removed from vaccines or that the efficacy needs to be better analyzed and that better understanding of serotype 3 could lead to a better incorporation of this serotype in a future vaccine formulation?
The central thesis of this paper is that the effectiveness of conjugate vaccines against serotype 3 is limited, evidence to support this statement is provided, along with evidence that the currently accepted correlate of protection is far too low. It is the opinion of the authors that, technical difficulties with the assay aside, the result for serotype 3 is of little to no clinical relevance as it is clearly established that production of antibody above the currently accepted correlate of protection does not necessarily confer immunity, and that negative results provide no information on the patient’s intrinsic ability to mount an antibody response given the poorly understood response to the vaccine.
We have expanded the discussion to make clearer the fact that evidence from the UK NEQAS showing poor reproducibility is somewhat incidental to the central thesis of the paper – the measurement itself is of no clinical value, and thus the effort required to address the evident technical issues is not worth pursuing.
We have expanded the explanation of the limited clinical usefulness of serotype 3 antibody measurements to make more explicit what was previously implicit in the discussion.
The authors recognise that suggesting the removal of a component of a vaccine already approved and in use would be of limited use, so long as that component is not actively harmful. We appreciate the suggestion that a better understanding of serotype 3 could lead to better incorporation in future vaccine formulation, and have added words to this effect.
Round 2
Reviewer 2 Report
The authors have greatly improved the clarity of manuscript.
Luminex assay:
In reading the revision, the authors now make it clear that these data come directly from the quality control programme data. For anyone not part of the network it is very difficult to know or find out how many laboratories are testing, etc even when going to the website. Adding a little more detail increases the appreciation of this communication to an audience of pneumococcal researchers that may not be part of the network but are interested in the overall discussion provided.If two samples are tested by all sites annually the data presented should in theory be derived from a 15-year period, for example.
suggest to add "annually" on line 245
When the authors state that the other serotypes generally exhibited a higher rate of agreement, what would that be greater than ? Can they be more precise? What would actually be acceptable?
Author Response
Luminex assay:
In reading the revision, the authors now make it clear that these data come directly from the quality control programme data. For anyone not part of the network it is very difficult to know or find out how many laboratories are testing, etc even when going to the website. Adding a little more detail increases the appreciation of this communication to an audience of pneumococcal researchers that may not be part of the network but are interested in the overall discussion provided.If two samples are tested by all sites annually the data presented should in theory be derived from a 15-year period, for example.
suggest to add "annually" on line 245
These samples are issued every two months, and we have data going back to 2016. The paper has been updated with this information. Numbers of labs have also been added - there are 10 labs participating with 6-8 results on any one occasion.
When the authors state that the other serotypes generally exhibited a higher rate of agreement, what would that be greater than ? Can they be more precise? What would actually be acceptable?
Other serotypes exhibit on average about 90% agreement. Of course, this is not a strictly fair comparison as concentrations may not be directly comparable for other serotypes, however this gives an approximate comparison with the ~65% agreement for the serotype 3 component. Of note, the other serotypes do not exhibit the effect seen with serotype 3 were agreement is paradoxically lower for higher GMC samples.
