To the editor—We read with interest the comments by Dr Musher about our manuscript entitled “High Frequency of Specific Polysaccharide Antibody Deficiency in Adults With Unexplained, Recurrent and/or Severe Infections With Encapsulated Bacteria” 1.Dr Musher notes that no etiologic diagnosis for the pneumonia was established in 23 of the 37 specific polysaccharide antibody deficiency (SPAD) patients. Indeed, patients were included when they had a history of recurrent and/or severe bacterial infections, and most were outcome patients with recurrent upper and/or lower respiratory tract infections (RTIs): bacterial identification is not easily available in such patients but must be encouraged. Dr Musher also noted that we did not confirm that Haemophilus infections were due to encapsulated organisms: once again, this is easily explained by the fact that this identification is not available in routine practice. However, the 3 patients had recurrent lower (n = 3) RTIs, upper RTIs (n = 1), and/or bronchiectasis (n = 2): documented Haemophilus infections in this context were just an additional information, and not the only reason for suspecting an antibody deficiency.Considering the extremely low frequency of meningococcal infections in healthy adults in absence of any outbreaks and in absence of any personal history of asplenia, human immunodeficiency virus (HIV) infection, active/passive smoking…, there is a broad consensus on the need to look for immune deficiencies such complement pathway deficiencies, but antibody deficiencies can also be found 2, 3.Dr Musher also adds that around 10% of healthy subjects can display low responses to pneumococcal capsular polysaccharides (PCP), implying that a poor response is not synonymous of immunodeficiency. We used the guidelines proposed by the American Academy of Allergy, Asthma and Immunology (AAAAI) and the American College of Allergy, Asthma and Immunology (ACAAI) for the use and the interpretation of anti-PCP response: these guidelines are widely accepted, despite the arbitrarily choices of the cutoff levels for anti-PCP antibodies 4–6. The most important is that we used these criteria in patients with relevant unexplained bacterial infections (and not healthy subjects). More robust arguments are needed to contest these expert recommendations.Dr Musher considers that we should have measured anti-PCP antibody concentration after pneumococcal conjugate vaccine (PCV) before suggesting immunoglobulin replacement therapy. Conjugate vaccines were given as the only treatment after SPAD diagnosis in a large part of patients with recurrent infections (10/23), when infections were not too frequent. However, a part of pneumococcal polysaccharide vaccine (PPV) non-responders (SPAD patients) are also PCV non responders suggesting that the defect of PCP recognition and/or specific antibody production is not bypassed by PCP conjugation with protein (6 and several personal unpublished observations in adult patients). We must also recall that the number of serotypes in available conjugate vaccines was limited to 13 and that some emerging non-PCV13 serotypes are responsible for most invasive infections in the general population 7: further studies are warranted to know if PCV containing 20 conjugated PCP or more, would change this practice.Dr Musher adds that “levels of antibody to pneumococcal capsular polysaccharide are relatively low in commercial preparations of IVIG”. We must reaffirm that all SPAD patients must not be treated with Ig replacement therapy. But we fully agree with some experts who consider immunoglobulin G (IgG) replacement therapy in SPAD patients who have severe or very frequent recurrent infections and/or bronchiectasis, inability to tolerate antibiotic prophylaxis (severe side effects or complications), or failure to respond to prophylactic antibiotics 8. Our study also supports this recommendation, since all patients were dramatically improved by Ig therapy: this confirms, if necessary, that the diagnosis criteria of this antibody deficiency seem appropriate in these patients, and that Ig preparations seem to contain enough anti-pneumococcal antibodies.