Reply

Abstract

A point-by-point reply to the specific criticisms by Cox et al1Cox L.S. Didier A. Demoly P. Wahn U. Pradalier A. Frew A.J. et al.Methodological aspects of a meta-analysis of grass pollen allergen sublingual immunotherapy tablets.J Allergy Clin Immunol. 2016; 138: 314-315.e4Abstract Full Text Full Text PDF PubMed Google Scholar of our meta-analysis2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (62) Google Scholar is available in this article's Online Repository at www.jacionline.org. Here we will focus on the most critical methodological defect of the sublingual immunotherapy (SLIT) randomized controlled trials (RCTs), which is the metric they used to assess the clinical benefit. This metric is mathematically incorrect because, as clearly explained in our work,2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (62) Google Scholar it calculates the percentage difference between SLIT and placebo, not taking into account the symptom score (SS) scale range and leading to a huge magnification of the difference between groups. By using this metric, a 1-point difference will be the same percentage difference in an 18-point scale (the most common SS scale used), a 100-point scale, or any other scale, and this is mathematically unacceptable (a detailed explanation has been reported in Fig 1 and our original article2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (62) Google Scholar). The correct metric, which takes into account the scale range, was indicated by the World Allergy Organization (WAO)3Canonica G.W. Baena-Cagnani C.E. Bousquet J. Bousquet P.J. Lockey R.F. Malling H.-J. et al.Recommendations for standardization of clinical trials with allergen specific immunotherapy for respiratory allergy. A statement of a World Allergy Organization (WAO) taskforce.Allergy. 2007; 62: 317-324Crossref PubMed Scopus (318) Google Scholar and is based on the comparison between the pretreatment and posttreatment SSs of the active and placebo groups. Using this metric in our work, we showed a small difference between SLIT and placebo, which is less than the US Food and Drug Administration (FDA; 15%) and WAO (20%) thresholds of efficacy.2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (62) Google Scholar The baseline in the case of SLIT RCTs is the retrospective (prior year) total symptom score (RTSS), which is used by the investigators of the original RCTs as inclusion criteria. In other words the RTSS is assumed by the investigators as the SS that the patients would have in the absence of any treatment (corresponding to the inclusion criteria). We acknowledge that the RTSS might be imprecise, but it should be similar to the SS of the treatment season, especially if the pollen count of the 2 consecutive seasons is similar, and we have shown for the Cox study (see the Methods section in this article's Online Repository) that this possible imprecision does not affect the results. In our work2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (62) Google Scholar we also reported the difference between SLIT and placebo not only in terms of the standardized mean difference (SMD) but also in terms of the mean difference (MD), which is the difference in SS points between SLIT and placebo. We showed that this difference is −0.83 SS points (95% CI, −1.03 to −0.63). In a recent work Devillier himself4Devillier P. Chassany O. Vicaut E. de Beaumont O. Robin B. Dreyfus J.F. et al.The minimally important difference in the Rhinoconjunctivitis Total Symptom Score in grass-pollen-induced allergic rhinoconjunctivitis.Allergy. 2014; 69: 1689-1695Crossref PubMed Scopus (20) Google Scholar estimated the minimally important difference, which is defined as the smallest improvement considered worthwhile by a patient, as 1.1 to 1.3 SS points in patients with grass pollen–related rhinoconjunctivitis. Therefore the difference of −0.83 SS points reported in our meta-analysis is less than the minimally important difference estimated by Devillier et al (see the Methods section in this article's Online Repository for details). In conclusion, the analyses based on the Devillier minimally important difference as the threshold of efficacy (1.1 SS points) confirms the conclusions of our work, estimating a small treatment benefit (less than the FDA's 15% or WAO's 20% difference thresholds) and that the incorrect metric used in the SLIT RCTs highly overstated the treatment benefit. We will reply point by point to the specific criticisms by Cox et alE1Cox L.S. Didier A. Demoly P. Wahn U. Pradalier A. Frew A.J. et al.Methodological aspects of a meta-analysis of grass pollen allergen sublingual immunotherapy tablets.J Allergy Clin Immunol. 2016; 138: 314-315.e4Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar of our meta-analysisE2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar and take advantage of their comments to better explain the clinical problems in the SLIT RCTs after a brief historical background. SLIT was approved by the FDA in 2014 but is not a new therapy; SLIT has been widely used in the United States for many years. Early literature includes many articles by a variety of authors.E3Hansel F.K. Sublingual testing and therapy. A comparison with injectable methods.Trans Am Soc Ophthalmol Otolaryngol Allergy. 1972; 11: 93-103Google Scholar, E4Dickey L.W. Sublingual antigen testing and therapy for inhalants, foods and petrochemicals.in: Dickey L. Clinical ecology. Charles C. Thomas, Springfield (IL)1976: 544-553Google Scholar, E5Pfeiffer G.O. Sublingual procedures.Trans Am Soc Ophthalmol Otolaryngol Allergy. 1970; 11: 104Google Scholar, E6Ruddy L.W. Why sublingual therapy in otolaryngologic allergy.Trans Am Soc Ophthalmol Otolaryngol Allergy. 1976; 17: 93-102Google Scholar, E7Waickman F.J. Pediatric allergy.Trans Am Soc Ophthalmol Otolaryngol Allergy. 1965; 6: 39Google Scholar, E8Hansel F.K. Allergy of the nose and paranasal sinuses. CV Mosby Co, St Louis (MO)1936: 755Google Scholar HanselE3Hansel F.K. Sublingual testing and therapy. A comparison with injectable methods.Trans Am Soc Ophthalmol Otolaryngol Allergy. 1972; 11: 93-103Google Scholar cites Black, who in 1928 reported on the successful management of pollen allergy by SLIT. A course called “Sublingual Therapy in Allergy” was offered at the American Academy of Otolaryngic Allergy from 1963 to 1980. In those early days, SLIT was available not only as drops but also as rapidly dissolving tablets that have only recently been introduced in Europe.E2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar, E9Kleine-Tebbe J. Ribel M. Herold D.A. Safety of a SQ-standardized grass allergen tablet for sublingual immunotherapy: a randomized, placebo-controlled trial.Allergy. 2006; 61: 181-184Crossref PubMed Scopus (116) Google Scholar The use of SLIT in the United States decreased markedly and almost “vanished” (an exception is the American Academy of Environmental Medicine, where Frank Waickman offered courses on sublingual treatments for years).E10Samter M. Sublingual desensitization for allergy not recommended. Questions and answer.JAMA. 1971; 215: 2120Google Scholar With the exception of a few publications by David Morris,E11Morris D.L. Treatment of respiratory disease with ultra-small doses of antigens.Ann Allergy. 1970; 28: 494-500PubMed Google Scholar no more articles about SLIT were published in the United States between 1970 and 1993, when Nelson et alE12Nelson H.S. Oppenheimer J. Vatsia G.A. Buchmeier A. A double-blind, placebo-controlled evaluation of sublingual immunotherapy with standardized cat extract.J Allergy Clin Immunol. 1993; 92: 229-236Abstract Full Text PDF PubMed Scopus (116) Google Scholar performed an RCT concluding that SLIT with high-dose standardized cat extracts was no more effective than placebo in reducing symptoms or affecting immunologic measures of cat sensitivity. An analysis not including the studies by Pradalier et al (100-IR tablets, a low dose compared with other studies using the 300-IR tablet)E13Pradalier A. Basset D. Claudel A. Couturier P. Wessel F. Galvain S. et al.Sublingual-swallow immunotherapy (SLIT) with a standardized five-grass-pollen extract (drops and sublingual tablets) versus placebo in seasonal rhinitis.Allergy. 1999; 54: 819-828Crossref PubMed Scopus (146) Google Scholar and Smith et al (3× 100-IR tablets administered 3 times per week instead of once daily)E14Smith H. White P. Annila I. Poole J. Andre C. Frew A. Randomized controlled trial of high-dose sublingual immunotherapy to treat seasonal allergic rhinitis.J Allergy Clin Immunol. 2004; 114: 831-837Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar is already present in the Results section of our studyE2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar using both the SMD and the MD as measures of efficacy: the SMD excluding these 2 studiesE13Pradalier A. Basset D. Claudel A. Couturier P. Wessel F. Galvain S. et al.Sublingual-swallow immunotherapy (SLIT) with a standardized five-grass-pollen extract (drops and sublingual tablets) versus placebo in seasonal rhinitis.Allergy. 1999; 54: 819-828Crossref PubMed Scopus (146) Google Scholar, E14Smith H. White P. Annila I. Poole J. Andre C. Frew A. Randomized controlled trial of high-dose sublingual immunotherapy to treat seasonal allergic rhinitis.J Allergy Clin Immunol. 2004; 114: 831-837Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar did not change (SMD, −0.28; 95% CI, −0.39 to −0.18; P < .0001), indicating that an SMD of −0.28 corresponds to a MD of −0.83 SS points.E2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar Therefore the exclusion of these 2 studies did not make any difference (same point estimate: SMD, −0.28 with or without the 2 studies), as already clearly pointed out in the text, and would have only led to a loss of data along with the information they provided. In contrast, it showed that the benefit does not change even if patients are treated with a low dose of grass allergens. Cox et alE1Cox L.S. Didier A. Demoly P. Wahn U. Pradalier A. Frew A.J. et al.Methodological aspects of a meta-analysis of grass pollen allergen sublingual immunotherapy tablets.J Allergy Clin Immunol. 2016; 138: 314-315.e4Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar are concerned about the conclusions of our work,E2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar mainly because of the metric we used to analyze the data. According to Cox et al,E1Cox L.S. Didier A. Demoly P. Wahn U. Pradalier A. Frew A.J. et al.Methodological aspects of a meta-analysis of grass pollen allergen sublingual immunotherapy tablets.J Allergy Clin Immunol. 2016; 138: 314-315.e4Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar this metric, based on the difference between the baseline SS, which is the prior year's (retrospective) SS, and the during-treatment SS cannot be used because (1) the prior year's SS might be different from that of the treatment season because of “many confounding factors,” without suggesting any (likely they mean the variability of pollen seasons), and (2) a recall bias might influence the reliability of the retrospective SS. The retrospective SS is also one of the main inclusion criteria (eg, SS >12 in the Cox study).E15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar Recall bias and the variability of pollen seasons can affect the retrospective SS, but it reflects the baseline value with an acceptable approximation, especially if a measurement of the pollen count in the treatment season (to be compared with that of the previous season) is provided. In any case, even if a recall bias was present and therefore the retrospective SS could have been lower than reported in the study (14.9), it did not affect the results. In fact, even if the real during-treatment SS in the pollen season was hypothetically 12 instead of 14.9 because of recall bias (about 3 points less than that reported in the study, which is the lower limit because the study enrolled patients with severe rhinoconjunctivitis according to the inclusion criteria, corresponding to an SS of >12), the percentage improvement would have been 8% (3.21-4.16 = −0.95/12 = −7.9%), which is close to the 6% calculated with the original data and still far from the FDA's 15% difference or the WAO's 20% difference threshold of minimum clinical efficacy. Therefore even admitting an imprecision of the baseline SS (because of confounding, like the variability of pollen season or recall bias), it does not significantly affect our calculation. Furthermore, recall bias could have influenced retrospective SSs in both directions, being even higher than that reported in the study by Cox et al (14.9).E15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar In any case we acknowledge that there is the theoretical possibility that an uncommonly low-level pollen season might cause symptoms that were less severe than expected in the study season compared with the previous season. However, it is very unlikely that this occurs for all the RCTs, and recall bias should affect the placebo and active groups similarly. Note that the metric we used, the same used for symptomatic drugs (eg, antihistamines or corticosteroids), was recommended by the WAO to be used for SLIT in the 2007 WAO position paper.E16Canonica G.W. Baena-Cagnani C.E. Bousquet J. Bousquet P.J. Lockey R.F. Malling H.J. et al.Recommendations for standardization of clinical trials with Allergen Specific Immunotherapy for respiratory allergy. A statement of a World Allergy Organization (WAO) taskforce.Allergy. 2007; 62: 317-324Crossref PubMed Scopus (380) Google Scholar This metric was also used to establish the 20% threshold of the “minimum relevant magnitude of efficacy” of SLIT. Therefore SLIT studies cannot use the threshold calculated with this metric and estimate the benefit by using a different metric, which is the difference in mean SS between the SLIT and placebo groups during the pollen season. Worse, the metric used in the RCTs is mathematically incorrect because it does not take into account the SS or medication score (MS) scale range, leading to a huge overestimation of the difference between treatments. In other words, using the metric of the RCTs, a 1-point difference between 2 treatments would be the same in a 10-point scale or a 100-point scale, as clearly shown in our study.E2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar Also, note that the main analysis of our studyE2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar estimated the treatment benefit with the incorrect metric used in the individual studies by means of the SMD, and the pooled estimate was −0.28 SMD (Fig 1, A, of our articleE2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar) for the SS (−0.24 excluding the influential study, see reference 16 in our articleE2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar). This is very close to the assumed threshold of inefficacy (−0.20 SMD), according to the Cohen criteria,E17Cohen J. Statistical power analysis for the behavioral sciences.2nd ed. Lawrence Erlbaum Associates, Hillsdale (NJ)1988Google Scholar indicating a small benefit even with the metric used in the SLIT RCTs. Therefore we do not cast doubt only on Cox et al's calculation of the RCI for the 5-grass pollen tabletsE15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar; this criticism is valid for all the SLIT grass RCTs, as well as for the SCIT studies. We cited the study by Cox et alE15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar as an example because it is the only one providing the retrospective SS, which can be used as the baseline SS to correct the mistaken calculation in the RCTs. In any case an analysis including all the SLIT RCTs using the correct metric (specified by WAO)E16Canonica G.W. Baena-Cagnani C.E. Bousquet J. Bousquet P.J. Lockey R.F. Malling H.J. et al.Recommendations for standardization of clinical trials with Allergen Specific Immunotherapy for respiratory allergy. A statement of a World Allergy Organization (WAO) taskforce.Allergy. 2007; 62: 317-324Crossref PubMed Scopus (380) Google Scholar is provided in our study,E2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar resulting in an average 6% difference, which is much lower than the WAO (20%) and FDA (15%) thresholds of efficacy. In conclusion, the approximation in the estimation of the baseline SS by using our metric is acceptable, does not affect the final results, and is mathematically correct in contrast to the metric used in the original RCTs. Furthermore, if it is valid to classify patients as having moderate or severe rhinoconjunctivitis (inclusion criteria) because we assume that this is the SS in the absence of any treatment it is valid as a baseline as well. These considerations are valid for the SCIT studies also, but unfortunately, they used different scoring systems, and a comparison can be made only through the SMD, according to the incorrect metric. However, by using the incorrect metric, an indirect comparison between SCIT and SLITE18Dretzke J. Meadows A. Novielli N. Huissoon A. Fry-Smith A. Meads C. Subcutaneous and sublingual immunotherapy for seasonal allergic rhinitis: a systematic review and indirect comparison.J Allergy Clin Immunol. 2013; 131: 1361-1366Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar, E19Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of subcutaneous and sublingual immunotherapy with grass allergens for seasonal allergic rhinitis: a meta-analysis-based comparison.J Allergy Clin Immunol. 2012; 130: 1097-1107.e2Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar estimated SCIT's efficacy as twiceE18Dretzke J. Meadows A. Novielli N. Huissoon A. Fry-Smith A. Meads C. Subcutaneous and sublingual immunotherapy for seasonal allergic rhinitis: a systematic review and indirect comparison.J Allergy Clin Immunol. 2013; 131: 1361-1366Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar or tripleE19Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of subcutaneous and sublingual immunotherapy with grass allergens for seasonal allergic rhinitis: a meta-analysis-based comparison.J Allergy Clin Immunol. 2012; 130: 1097-1107.e2Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar that of SLIT. Based on the considerations reported above about the correct metric to be used, it is clear that all the analyses shown in the Matricardi et al,E20Matricardi P.M. Kuna P. Panetta V. Wahn U. Narkus A. Subcutaneous immunotherapy and pharmacotherapy in seasonal allergic rhinitis: a comparison based on meta-analyses.J Allergy Clin Immunol. 2011; 128: 791-799Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar Devillier et al,E21Devillier P. Dreyfus J.F. Demoly P. Calderón M.A. A meta-analysis of sublingual allergen immunotherapy and pharmacotherapy in pollen-induced seasonal allergic rhinoconjunctivitis.BMC Med. 2014; 12: 71Crossref PubMed Scopus (75) Google Scholar and Howarth et alE22Howarth P. Malling H.J. Molimard M. Devillier P. Analysis of allergen immunotherapy studies shows increased clinical efficacy in highly symptomatic patients.Allergy. 2012; 67: 321-327Crossref PubMed Scopus (51) Google Scholar studies are flawed by the same mistaken calculations comparing the during-treatment differences but not the improvements in each treatment group. Furthermore, it must be emphasized that the comparisons are not between SLIT or SCIT efficacy versus pharmacotherapy (corticosteroids, leukotriene antagonists, or antihistamines) but between SLIT plus pharmacotherapy or SCIT plus pharmacotherapy during the pollen season versus pharmacotherapy alone during 14 days of maximum symptoms, making it difficult to clearly define the real efficacy of immunotherapy alone. Regarding the appropriateness of analyzing SSs and MSs separately but not as a combined score, we want to emphasize that this is precisely the most critical problem in the SLIT RCTs. The total combined score (TCS), as clearly discussed in our article, is mathematically incorrect because the SS and MS scales have different weights depending on their range. The only study of the 13 that takes into account this aspect is the study by Cox et al,E15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar in which an MS with a comparable scale (0-3) was used and the TCS was normalized to obtain a correct result. Moreover, a standardization for the MS, in contrast to the SS, has never been done, with different points assigned to the same drugs in a totally arbitrary fashion. For example, oral or ocular antihistamine drops are equal to 1 point in the study by Cox et alE15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar and up to 6 points in the score used by Nelson et al.E23Nelson H.S. Nolte H. Creticos P. Maloney J. Wu J. Bernstein D.I. Efficacy and safety of timothy grass allergy immunotherapy tablet treatment in North American adults.J Allergy Clin Immunol. 2011; 127 (e1-2): 72-80Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar The same is true for corticosteroids: 2 or 3 points in the study by Cox et alE15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar and 8 to 16 points in the study by Nelson et al.E23Nelson H.S. Nolte H. Creticos P. Maloney J. Wu J. Bernstein D.I. Efficacy and safety of timothy grass allergy immunotherapy tablet treatment in North American adults.J Allergy Clin Immunol. 2011; 127 (e1-2): 72-80Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar However, the problem is more serious from a statistical point of view: as already pointed out in our article, the MS is a qualitative variable; it can be considered on an ordinal scale but cannot be treated as interval data because intervals between each value (ie, antihistamines = 1 point, nasal steroids = 2 points, and oral steroids = 3 points arbitrarily assigned) are not equal. Thus a TCS, even considering the adjustment proposed in the study by Cox et al,E15Cox L.S. Casale T.B. Nayak A.S. Bernstein D.I. Creticos P.S. Ambroisine L. et al.Clinical efficacy of 300IR 5-grass pollen sublingual tablet in a US study: the importance of allergen-specific serum IgE.J Allergy Clin Immunol. 2012; 130: 1327-1334.e1Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar or a standardization between the different MS scales cannot be computed, and scientific societies and regulatory agencies should confront this problem. If an adjustment must be made in SSs, it must be mathematically correct, rather than using a measure, such as the TCS (SS+MS), which only leads to a misleading magnification of a small SLIT benefit, as shown by the small SMD for SSs and MSs and the small percentage difference between SLIT and placebos SSs (6%, on average, which is far less than the FDA's 15% threshold of minimum clinical efficacy) obtained by using our correct metric, as specified by the WAO.E16Canonica G.W. Baena-Cagnani C.E. Bousquet J. Bousquet P.J. Lockey R.F. Malling H.J. et al.Recommendations for standardization of clinical trials with Allergen Specific Immunotherapy for respiratory allergy. A statement of a World Allergy Organization (WAO) taskforce.Allergy. 2007; 62: 317-324Crossref PubMed Scopus (380) Google Scholar We reported an average difference between SLIT and placebo of −0.83 SS points, which is less than 1 point. Devillier at alE24Devillier P. Chassany O. Vicaut E. de Beaumont O. Robin B. Dreyfus J.F. et al.The minimally important difference in the Rhinoconjunctivitis Total Symptom Score in grass-pollen-induced allergic rhinoconjunctivitis.Allergy. 2014; 69: 1689-1695Crossref PubMed Scopus (31) Google Scholar estimated the minimally important difference as a 1.1- to 1.3-point SS improvement; therefore according to the Devillier estimation,E24Devillier P. Chassany O. Vicaut E. de Beaumont O. Robin B. Dreyfus J.F. et al.The minimally important difference in the Rhinoconjunctivitis Total Symptom Score in grass-pollen-induced allergic rhinoconjunctivitis.Allergy. 2014; 69: 1689-1695Crossref PubMed Scopus (31) Google Scholar the difference we reported (−0.83 SS points) between SLIT and placebo is not perceived by the patients as clinically important, confirming our conclusions. Fig 1, B, of our workE2Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of grass pollen allergen sublingual immunotherapy tablets for seasonal allergic rhinoconjunctivitis: a systematic review and meta-analysis.JAMA Intern Med. 2015; 175: 1301-1309Crossref PubMed Scopus (77) Google Scholar shows the 95% CI for the mean value (−1.03 to −0.63 for the random effect and even smaller for the fixed effect). This implies that we are 95% confident that this interval contains the true value of the parameter. Therefore −1.03 could be a value for the population parameter: even if it was the true value (the most favorable extreme to SLIT), the probability of observing a value of less than −1.1 is only 0.25 (25% of patients could benefit significantly from SLIT). In contrast, if the true value was that reported in our study as a point estimate, less than 0.5% of patients can show an improvement of greater than −1.1. This is in accordance with the calculation using our metric reporting an SS reduction to less than the WAO (20%) and FDA (15%) thresholds of efficacy. Regarding safety, we agree with Cox et al that the majority of adverse events are mild to moderate and that “both SCIT and SLIT are very safe”,E25Aasbjerg K. Dalhoff K.P. Backer V. Adverse events during immunotherapy against grass pollen-induced allergic rhinitis—differences between subcutaneous and sublingual treatment.Basic Clin Pharmacol Toxicol. 2015; 117: 73-84Crossref PubMed Scopus (12) Google Scholar but as we showed in our previous meta-analysisE19Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of subcutaneous and sublingual immunotherapy with grass allergens for seasonal allergic rhinitis: a meta-analysis-based comparison.J Allergy Clin Immunol. 2012; 130: 1097-1107.e2Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar indirectly comparing SCIT and SLIT, the withdrawal rate for adverse events was higher in the SLIT group (78 patients; 0.04% vs 0.013% in the placebo group) than in the SCIT group (18 patients; 0.019% vs 0.005% in the placebo group).E19Di Bona D. Plaia A. Leto-Barone M.S. La Piana S. Di Lorenzo G. Efficacy of subcutaneous and sublingual immunotherapy with grass allergens for seasonal allergic rhinitis: a meta-analysis-based comparison.J Allergy Clin Immunol. 2012; 130: 1097-1107.e2Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar This evidence should also be considered. Regarding the criticism that our review does not take into account the long-term benefit of the treatment, it should be emphasized that the long-term benefit was not the focus of this review nor of the study by Cox et al nor of all the other RCTs. Regarding the physician-patient dialog to respect patient preference according to evidence-based medicine principles,E26Sackett D.L. Rosenberg W.M. Gray J.A. Haynes R.B. Richardson W.S. Evidence based medicine: what it is and what it isn't.BMJ. 1996; 312: 71-72Crossref PubMed Google Scholar, E27Haynes R.B. Devereaux P.J. Guyatt G.H. Physicians' and patients' choices in evidence based practice.BMJ. 2002; 324: 1350Crossref PubMed Google Scholar we believe that in the case of SLIT, the patient has to be informed correctly about the small benefit of the treatment. In the interest of patients, caution must be exercised when such a small treatment benefit is reported, especially if one considers that sponsored studies (as in the case of all SLIT RCTs) always show greater benefit compared with independent studies using the same drugs or devices.E28Lundh A. Sismondo S. Lexchin J. Busuioc O.A. Bero L. Industry sponsorship and research outcome.Cochrane Database Syst Rev. 2012; 12: MR000033Crossref PubMed Scopus (803) Google Scholar, E29Drugs and devices look more effective in studies sponsored by industry.BMJ. 2012; 345: e8386Crossref PubMed Google Scholar Methodological aspects of a meta-analysis of grass pollen allergen sublingual immunotherapy tabletsJournal of Allergy and Clinical ImmunologyVol. 138Issue 1PreviewWe have identified a number of issues affecting the conclusions of a recently published systematic review and meta-analysis of the efficacy of grass pollen allergen sublingual immunotherapy (SLIT) tablets for seasonal allergic rhinoconjunctivitis.1 The validity of a meta-analysis depends on the quality of the systematic review on which it is based and, to a lesser extent, on the conclusions that the authors draw from it. Although Di Bona et al1 applied a rigorous methodology to the meta-analysis per se, many aspects of their report, including their interpretation of the results, require clarification and comments acutely described in this article's Online Repository at www.jacionline.org . Full-Text PDF