Selective IgA deficiency in humans is associated with reduced gut microbial diversity
2019; Elsevier BV; Volume: 143; Issue: 5 Linguagem: Inglês
10.1016/j.jaci.2019.01.019
ISSN1097-6825
AutoresSilje F. Jørgensen, Kristian Holm, Magnhild E. Macpherson, Christopher Storm-Larsen, Martin Kummen, Børre Fevang, Pål Aukrust, Johannes R. Hov,
Tópico(s)Pediatric health and respiratory diseases
ResumoSelective IgA deficiency (SIgAD) is the most common primary immunodeficiency affecting 1:700 Caucasians. Many patients are asymptomatic, but SIgAD is associated with increased risk of autoimmune disease, gastrointestinal disorders, and bacterial respiratory tract infections in some individuals.1Yazdani R. Azizi G. Abolhassani H. Aghamohammadi A. Selective IgA deficiency: epidemiology, pathogenesis, clinical phenotype, diagnosis, prognosis and management.Scand J Immunol. 2017; 85: 3-12Crossref PubMed Scopus (109) Google Scholar In the gut, IgA is important for mucosal defense against microbes and retaining a mutual beneficial interaction with commensal bacteria, favoring microbes beneficial to a healthy gut environment.2Kaetzel C.S. Cooperativity among secretory IgA, the polymeric immunoglobulin receptor, and the gut microbiota promotes host-microbial mutualism.Immunol Lett. 2014; 162: 10-21Crossref PubMed Scopus (109) Google Scholar Murine models have shown that absence of IgA or impaired IgA production affects the balance of the gut microbiota, with increased permeability and influx of microbial components leading to increased systemic inflammation.3Kato L.M. Kawamoto S. Maruya M. Fagarasan S. The role of the adaptive immune system in regulation of gut microbiota.Immunol Rev. 2014; 260: 67-75Crossref PubMed Scopus (85) Google Scholar We recently reported that patients with common variable immunodeficiency (CVID) have reduced microbial diversity, which was more pronounced in patients with undetectable IgA in serum (IgA < 0.1 g/L),4Jorgensen S.F. Troseid M. Kummen M. Anmarkrud J.A. Michelsen A.E. Osnes L.T. et al.Altered gut microbiota profile in common variable immunodeficiency associates with levels of lipopolysaccharide and markers of systemic immune activation.Mucosal Immunol. 2016; 9: 1455-1465Crossref PubMed Scopus (36) Google Scholar suggesting that the absence of IgA affects gut microbial composition also in humans. We therefore hypothesized that patients with SIgAD have altered microbial diversity compared with controls. All adult patients registered with the International Classification of Diseases, Tenth Revision code of SIgAD (D80.2) between 2003 and 2015 at the Section of Clinical Immunology and Infectious Diseases, Oslo University Hospital were prescreened for eligibility (n = 83). Primary SIgAD was defined by serum IgA level of less than 0.7 g/L and exclusion of other causes of hypogammaglobulinemia. In the end, 31 patients with SIgAD were eligible and 15 patients sent in a stool sample. One of these later regained normal IgA levels and was therefore excluded (see Fig E1 in this article's Online Repository at www.jacionline.org). Healthy controls were recruited from the Norwegian Bone Marrow Registry, and patients with CVID were disease controls. Exclusion criteria for all participants were no antibiotics in the last month (including prophylactic antibiotics), sample older than 72 hours, colostomy, ileostomy, and nasogastric feeding tube. The study was approved by the Regional Committee for Medical and Health Research Ethics of South-Eastern Norway. All participants provided written, informed consent. Sample collection, DNA extraction, sequencing (16S rRNA, V3-V4), postsequencing processing, and operational taxonomic unit mapping to the Silva database is described in this article's Methods section in the Online Repository at www.jacionline.org. Analysis of composition of microbiomes5Mandal S. Van Treuren W. White R.A. Eggesbo M. Knight R. Peddada S.D. Analysis of composition of microbiomes: a novel method for studying microbial composition.Microb Ecol Health Dis. 2015; 26: 27663Crossref PubMed Google Scholar was used to identify taxa at genus level that were differently abundant between patients with SIgAD, patients with CVID, and healthy controls. Analysis of composition of microbiomes uses compositional log-ratios to compare mean abundance of the microbial taxa with each other, making no distributional assumptions, both improving power and false-discovery rates, compared with other methods.5Mandal S. Van Treuren W. White R.A. Eggesbo M. Knight R. Peddada S.D. Analysis of composition of microbiomes: a novel method for studying microbial composition.Microb Ecol Health Dis. 2015; 26: 27663Crossref PubMed Google Scholar After quality controls, stool samples from 14 patients with SIgAD, 50 patients with CVID, and 166 controls were analyzed (see Figs E1 and E2 and Table E1, Table E2, Table E3 in this article's Online Repository at www.jacionline.org). All patients and controls are Caucasian and except for 3 patients with CVID (oral budesonide, n = 2; prednisolone 5 mg, n = 1), no patients/controls were using immunosuppressive medication at inclusion. Microbial diversity (Chao1, a bacterial richness estimate) was lower in patients with SIgAD than in healthy controls (P < .001; Fig 1, A), but similar in patients with SIgAD and CVID. Also, other alpha diversity metrics such as Shannon (P = .005) and Faith's PD (P = .004) showed a significant reduction in alpha diversity in patients with SIgAD than in controls. We found no association between the "number of courses of antibiotics in the last year" and alpha diversity (Fig 1, B). In addition, we applied linear regression using Chao1 as a dependent variable and "disease phenotype" (ie, SIgAD vs healthy controls) together with "sex," "age," "body mass index," and "number of courses of antibiotics in the last year" as independent variables. Disease phenotype was the strongest predictor of Chao1 (β, −0.22; P = .008) followed by age (β, 0.16; P = .016) and sex (β, −0.15; P = .045). Number of courses of antibiotics in the last year and body mass index were not significant predictors (β, −0.12, P = .163, and β, −0.10, P = .190, respectively). Neither diet (vegetarian, n = 2; lactose-free diet, n = 2) nor clinical phenotype (recurrent infections, n = 11; autoimmunity, n = 4; enteropathy, n = 1; asymptomatic, n = 2) appeared to determine the alpha diversity in SIgAD, but the numbers were too small for statistics (see Fig E3, A and B, and Table E2 in this article's Online Repository at www.jacionline.org). Twenty-one taxa were significantly different in patients with SIgAD compared with healthy controls (Table I). Particularly, Phascolarctobacterium, Flavonifractor, Intestinimonas and a not specified Eubacterium from the fissicatena group in the Lachnospiraceae family were increased in patients with SIgAD, whereas 3 different taxa from the Ruminococcaceae, Lachnospiraceae, and Erysipelotrichaceae family were decreased in patients with SIgAD compared with controls.Table ISignificantly different genus between patients with SIgAD and healthy controls using Analysis of composition of microbiomesTaxaPercentile for healthy controlsPercentile for patients with SIgAD02550751000255075100Actinobacteria.Coriobacteriia.Coriobacteriales.Coriobacteriaceae.Adlercreutzia11135911111Cyanobacteria.Melainabacteria.Gastranaerophilales.gut metagenome.gut metagenome111116311111Firmicutes.Clostridia.Clostridiales.Lachnospiraceae.[Eubacterium] fissicatena group∗The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in patients with SIgAD compared with healthy controls.1111117121589Firmicutes.Clostridia.Clostridiales.Lachnospiraceae.Lachnospiraceae AC2044 group11138311111Firmicutes.Clostridia.Clostridiales.Lachnospiraceae.uncultured bacterium†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in healthy controls compared with patients with SIgAD.112717111111Firmicutes.Clostridia.Clostridiales.Peptococcaceae.Peptococcus111178811111Firmicutes.Clostridia.Clostridiales.Ruminococcaceae.Candidatus Soleaferrea11112011111Firmicutes.Clostridia.Clostridiales.Ruminococcaceae.Flavonifractor∗The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in patients with SIgAD compared with healthy controls.1614431,985231903048173,628Firmicutes.Clostridia.Clostridiales.Ruminococcaceae.Intestinimonas∗The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in patients with SIgAD compared with healthy controls.1381610819961212508,507Firmicutes.Clostridia.Clostridiales.Ruminococcaceae.Oscillospira111434911111Firmicutes.Clostridia.Clostridiales.Ruminococcaceae.Ruminococcaceae UCG-009†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in healthy controls compared with patients with SIgAD.113910911111Firmicutes.Erysipelotrichia.Erysipelotrichales.Erysipelotrichaceae.Dielma11111811111Firmicutes.Erysipelotrichia.Erysipelotrichales.Erysipelotrichaceae.uncultured†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in healthy controls compared with patients with SIgAD.1121415,97811111Firmicutes.Negativicutes.Selenomonadales.Acidaminococcaceae.Phascolarctobacterium∗The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in patients with SIgAD compared with healthy controls.11311,19322,873111,47127,07751,71980,868Lentisphaerae.Lentisphaeria.Victivallales.vadinBE97.uncultured bacterium111118811111Lentisphaerae.Lentisphaeria.Victivallales.Victivallaceae.Victivallis111414611111Proteobacteria.Betaproteobacteria.Burkholderiales.Oxalobacteraceae.Oxalobacter11155911111Proteobacteria.Gammaproteobacteria.Alteromonadales.Shewanellaceae.Shewanella1111311111Proteobacteria.Gammaproteobacteria.Oceanospirillales.Halomonadaceae.Halomonas1111711111Tenericutes.Mollicutes.Mollicutes RF9.uncultured bacterium.uncultured bacterium111120611111Only those taxa that were present in at least 5 samples were included in analysis. For each significantly different taxon, the number of reads per phyla is given in percentiles for healthy controls and patients with SIgAD. For example, 50th percentile for Flavonifractor means that in half the samples from patients with SIgAD, 304 or fewer sequences were assigned to Flavonifractor compared with only 14 in healthy controls.∗ The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in patients with SIgAD compared with healthy controls.† The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) increased in healthy controls compared with patients with SIgAD. Open table in a new tab Only those taxa that were present in at least 5 samples were included in analysis. For each significantly different taxon, the number of reads per phyla is given in percentiles for healthy controls and patients with SIgAD. For example, 50th percentile for Flavonifractor means that in half the samples from patients with SIgAD, 304 or fewer sequences were assigned to Flavonifractor compared with only 14 in healthy controls. The CVID cohort included 18 patients (and 166 controls) from the previous published cohort4Jorgensen S.F. Troseid M. Kummen M. Anmarkrud J.A. Michelsen A.E. Osnes L.T. et al.Altered gut microbiota profile in common variable immunodeficiency associates with levels of lipopolysaccharide and markers of systemic immune activation.Mucosal Immunol. 2016; 9: 1455-1465Crossref PubMed Scopus (36) Google Scholar and baseline samples from 32 patients with CVID in a recently published intervention study6Jorgensen S.F. Macpherson M.E. Bjørnetrø T. Holm K. Kummen M. Rashidi A. et al.Rifaximin alters gut microbiota profile, but does not affect systemic inflammation—a randomized controlled trial in common variable immunodeficiency.Sci Rep. 2019; 9: 167Google Scholar (Fig E2). In the present sample set, we found different abundance of bacterial taxa and a lower alpha diversity in patients with CVID than in controls (Fig 1, A; see Table E4 in this article's Online Repository at www.jacionline.org), as previously described.4Jorgensen S.F. Troseid M. Kummen M. Anmarkrud J.A. Michelsen A.E. Osnes L.T. et al.Altered gut microbiota profile in common variable immunodeficiency associates with levels of lipopolysaccharide and markers of systemic immune activation.Mucosal Immunol. 2016; 9: 1455-1465Crossref PubMed Scopus (36) Google Scholar The low alpha diversity was independent of the presence of enteropathy in patients with CVID (P = .35; see Fig E4 in this article's Online Repository at www.jacionline.org) and those 3 who used low-dose immunosuppressive drugs did not differ from the other patients with CVID. We have previously found significantly lower alpha diversity in patients with CVID with undetectable IgA in serum (IgA < 0.1 g/L) compared with patients with CVID with low to normal IgA in serum (IgA ≥ 0.1 g/L).4Jorgensen S.F. Troseid M. Kummen M. Anmarkrud J.A. Michelsen A.E. Osnes L.T. et al.Altered gut microbiota profile in common variable immunodeficiency associates with levels of lipopolysaccharide and markers of systemic immune activation.Mucosal Immunol. 2016; 9: 1455-1465Crossref PubMed Scopus (36) Google Scholar Herein, however, this finding was not confirmed (P = .80; see Fig E5 in this article's Online Repository at www.jacionline.org). One possible explanation is that in both studies most (98%4Jorgensen S.F. Troseid M. Kummen M. Anmarkrud J.A. Michelsen A.E. Osnes L.T. et al.Altered gut microbiota profile in common variable immunodeficiency associates with levels of lipopolysaccharide and markers of systemic immune activation.Mucosal Immunol. 2016; 9: 1455-1465Crossref PubMed Scopus (36) Google Scholar and 94% [present study]) patients with CVID had reduced IgA (IgA < 0.7 g/L) in serum (see Table E5 in this article's Online Repository at www.jacionline.org) and that the cutoff of IgA less than 0.1 is too strict; that is, reduced IgA may be equally important as not measurable IgA levels in serum. Also, a very recent publication showed that reduced IgA expression in duodenal biopsies was linked to the subgroup of patients with CVID with enteropathy. These patients also had increased occurrence of certain bacteria (ie, Acinetobacter baumannii) in the duodenal microbial community that induced an inflammatory phenotype in macrophages.7Shulzhenko N. Dong X. Vyshenska D. Greer R.L. Gurung M. Vasquez-Perez S. et al.CVID enteropathy is characterized by exceeding low mucosal IgA levels and interferon-driven inflammation possibly related to the presence of a pathobiont.Clin Immunol. 2018; 197: 139-153Crossref PubMed Scopus (34) Google Scholar These findings further support a role of IgA in CVID and a possible link to systemic inflammation. The present data shed new light on the impact of IgA on the gut microbiota composition in humans. We found a strong association between reduced intraindividual diversity and SIgAD, and given the presence of reduced diversity in mouse models of IgA deficiencies,3Kato L.M. Kawamoto S. Maruya M. Fagarasan S. The role of the adaptive immune system in regulation of gut microbiota.Immunol Rev. 2014; 260: 67-75Crossref PubMed Scopus (85) Google Scholar it is tempting to speculate that shared mechanisms explain these observations. In the only previous study of SIgAD, investigating 17 patients and 34 controls, microbial diversity was similar in the 2 groups.8Fadlallah J. El Kafsi H. Sterlin D. Juste C. Parizot C. Dorgham K. et al.Microbial ecology perturbation in human IgA deficiency.Sci Transl Med. 2018; 10: eaan1217Crossref PubMed Scopus (157) Google Scholar Shotgun sequencing as used in that study can differ from 16S-based methods,9Tessler M. Neumann J.S. Afshinnekoo E. Pineda M. Hersch R. Velho L.F.M. et al.Large-scale differences in microbial biodiversity discovery between 16S amplicon and shotgun sequencing.Sci Rep. 2017; 7: 6589Crossref PubMed Scopus (118) Google Scholar but it is unclear whether that explain the lack of differences between the groups. In the previous study, the main taxa found upregulated and downregulated in IgA were mostly from the Firmicutes phylum. This was in line with the present results, but because of differences in methods, the data were not comparable at the genus level. However, in the study by Fadlallah et al,8Fadlallah J. El Kafsi H. Sterlin D. Juste C. Parizot C. Dorgham K. et al.Microbial ecology perturbation in human IgA deficiency.Sci Transl Med. 2018; 10: eaan1217Crossref PubMed Scopus (157) Google Scholar the patients with SIgAD were younger and leaner, had less respiratory tract infections, had more autoimmune disease, and 14% used immunosuppressive drugs that was an exclusion criterion in the present study and all these factors could clearly influence the gut microbiota. Particularly, the difference in autoimmunity and recurrent infections points to possible explanatory and confounding factors that influence the gut microbiota in human immunoglobulin deficiencies. Although there was no association with "antibiotics use in the last year" and alpha diversity in our study, we cannot exclude the long-term effects of previous antibiotics use, or effects that are more subtle than can be detected in a study with our sample size and sequencing depth. Of note, both studies recruited patients from a hospital setting and are not representative of a general SIgAD population in terms of symptoms/phenotypes, and we consider this the main weakness of both studies. A next step to understand the true biological effects of IgA will be to (1) include gut microbiota data from a large number of asymptomatic SIgAD individuals (identified by screening in population cohorts), (2) explore the role of secretory IgA/IgM of the gut/oral cavity, (3) include data on inflammatory markers of systemic inflammation and endotoxin, and (4) apply interventional "proof-of-concept" study designs targeting the gut in humans with SIgAD and simultaneously assess effects on the gut microbiome, systemic inflammation, and immunity. This could potentially identify a link between gut microbial diversity, systemic inflammation, and IgA phenotype (eg, autoimmunity) also in humans. In conclusion, we found reduced microbial diversity in patients with SIgAD that parallels the findings in murine IgA deficiency, in addition to altered abundances of multiple bacterial taxa. However, the number of patients with SIgAD in the study is low and larger studies are needed to make a firm conclusion on the role of gut microbiota in this primary immunodeficiency. Reads containing Illumina Universal Adapters or PhiX were discarded using bbduk version 37.75 (BBTools; https://jgi.doe.gov/data-and-tools/bbtools) (parameters adapters: k = 23 hdist = 1 tbo cf = TRUE ftm = 5. parameters phix: k = 31 hdist = 1) and the remaining reads were demultiplexed using Je version 1.2E1Girardot C. Scholtalbers J. Sauer S. Su S.-Y. Furlong E.E. Je, a versatile suite to handle multiplexed NGS libraries with unique molecular identifiers.BMC Bioinform. 2016; 17: 419Crossref PubMed Scopus (76) Google Scholar (parameters: MAX_MISMATCHES = 1 MIN_MISMATCH_DELTA = 2). Indexes, heterogeneity spacers, and primers were trimmed with cutadapt version 1.14E2Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads.EMBnet J. 2011; 17: 10-12Crossref Google Scholar (parameter: --overlap 20 --discard-untrimmed -m 200) and the paired end reads were subsequently quality trimmed and merged using bbmerge version 37.75E3Bushnell B. Rood J. Singer E. BBMerge–accurate paired shotgun read merging via overlap.PloS One. 2017; 12: e0185056Crossref PubMed Scopus (463) Google Scholar (parameters: qtrim = rl trimq = 15 maxlength = 440 mininsert = 390 --loose 2). The merged contigs were quality filtered using default values in Quantitative Insights Into Microbial Ecology (QIIME) version 1.9.1.E4Caporaso J.G. Kuczynski J. Stombaugh J. Bittinger K. Bushman F.D. Costello E.K. et al.QIIME allows analysis of high-throughput community sequencing data.Nat Methods. 2010; 7: 335-336Crossref PubMed Scopus (24741) Google Scholar Closed reference operational taxonomic unit (OTU) mapping to the Silva databaseE5Quast C. Pruesse E. Yilmaz P. Gerken J. Schweer T. Yarza P. et al.The SILVA ribosomal RNA gene database project: improved data processing and web-based tools.Nucleic Acids Res. 2012; 41: D590-D596Crossref PubMed Scopus (14000) Google Scholar (version 128, reference OTUs preclustered at 97% sequence similarity) was performed using SortMeRNA version 2.0 through QIIME.E6Kopylova E. Noé L. Touzet H. SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data.Bioinformatics. 2012; 28: 3211-3217Crossref PubMed Scopus (1246) Google Scholar OTUs with a number of sequences less than 0.005% of the total number of mapped sequences were discarded as recommended.E7Bokulich N.A. Subramanian S. Faith J.J. Gevers D. Gordon J.I. Knight R. et al.Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing.Nat Meth. 2013; 10: 57Crossref PubMed Scopus (2509) Google Scholar The samples were rarefied (subsampled) to an OTU count of 8144 per sample, and all further analyses (except differential abundance with AncomE8Mandal S. Van Treuren W. White R.A. Eggesbo M. Knight R. Peddada S.D. Analysis of composition of microbiomes: a novel method for studying microbial composition.Microb Ecol Health Dis. 2015; 26: 27663Crossref PubMed Google Scholar) were performed on this rarefied data set. Calculation of alpha diversity and analysis of differential taxonomic composition were done in QIIME version 2018.4. The CVID samples were recruited from 2 previous published studiesE9Jorgensen S.F. Troseid M. Kummen M. Anmarkrud J.A. Michelsen A.E. Osnes L.T. et al.Altered gut microbiota profile in common variable immunodeficiency associates with levels of lipopolysaccharide and markers of systemic immune activation.Mucosal Immunol. 2016; 9: 1455-1465Crossref PubMed Scopus (88) Google Scholar, E10Jorgensen S.F. Macpherson M.E. Bjørnetrø T. Holm K. Kummen M. Rashidi A. et al.Rifaximin alters gut microbiota profile, but does not affect systemic inflammation—a randomized controlled trial in common variable immunodeficiency.Sci Rep. 2019; 9: 167Google Scholar (n = 84). From the intervention study,E10Jorgensen S.F. Macpherson M.E. Bjørnetrø T. Holm K. Kummen M. Rashidi A. et al.Rifaximin alters gut microbiota profile, but does not affect systemic inflammation—a randomized controlled trial in common variable immunodeficiency.Sci Rep. 2019; 9: 167Google Scholar all baseline samples (n = 40) were included because these CVID samples had not previously been compared with healthy controls. In the cross-sectional study (n = 44), 17 patients were excluded because the individuals were also included in the intervention study. In the end, 67 CVID samples were included for sequencing, and of these 17 samples were excluded after failing quality control and/or OTU count of less than 8144 (see above), resulting in samples from 50 patients with CVID as disease controls. Fig E3 shows the inclusion/exclusion of the CVID control group. Because this disease control group consists of 2 CVID cohorts, there are 2 minor differences in exclusion/inclusion criteria compared with the patients with SIgAD. First, patients with CVID who were vegetarians or vegans were excluded, whereas this was not an exclusion criterion in the SIgAD study. Second, all immunosuppressive drugs were an exclusion criterion in the intervention study,E10Jorgensen S.F. Macpherson M.E. Bjørnetrø T. Holm K. Kummen M. Rashidi A. et al.Rifaximin alters gut microbiota profile, but does not affect systemic inflammation—a randomized controlled trial in common variable immunodeficiency.Sci Rep. 2019; 9: 167Google Scholar whereas the cross-sectional study had all immunosuppressive drugs apart from 5 mg prednisolone or lower (or equivalent systemic dose of corticoid steroids) as an exclusion criterion.Fig E2Flowchart of inclusion/exclusion of patients with CVID in the study.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3Alpha diversity (Chao1) in SIgAD for different diets (A) and clinical phenotype (B).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E4Alpha diversity (Chao1) in CVID separated by enteropathy. Alpha diversity (Chao1) comparing CVID with and without enteropathy. Shown in median and quartiles. The comparisons are made by Mann-Whitney U test.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E5Alpha diversity (Chao1) in CVID separated by IgA levels. Alpha diversity (Chao1) comparing CVID with serum IgA levels of less than 0.1 g/L versus CVID with IgA levels of greater than or equal to 0.1 g/L. Shown in median, quartiles, and range. The comparisons are made by Mann-Whitney U test.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table E1Overview of the characteristics for patients with SIgAD, controls, and patients with CVIDCharacteristicSIgAD (n = 14)CVID (n = 50)Controls (n = 166)P valueSIgAD vs controlsAll phenotypesAge (y), median ± SD (range)41 ± 15 (19-68)48 ± 14 (18-82)47 ± 7 (30-59).51.16Sex: male, %364863.08.04Body mass index, median ± SD26 ± 425 ± 426 ± 4.38.06Duration of disease, median ± SD (range)10 ± 5 (4-16)16 ± 11 (0-53)———Recurrent infections, n (%)11 (79)50 (100)———Autoimmunity, n (%)4 (29)20 (40)———IgA < 0.1, n (%)13 (93)36 (72)———Enteropathy,∗Defined as diarrhea for at least 3 months with exclusion of infectious cause at the time of inclusion. Based on our recent study,E11 this was not necessarily associated with immune-mediated pathology on biopsy specimens. n (%)1 (7)14 (28)†Three additional patients with CVID had enteropathy, but they were excluded because of the use of immunosuppressive therapy.———No. of courses of antibiotics in the last year,‡Excluding the last month. median ± SD (range)1.5 ± 0.6 (0-8)2 ± 2.3 (0-7)0 ± 2.1 (0-3)9.1 × 10−76.0 × 10−15Median IgG level (g/mL), median ± SD13.0 ± 2.0 (10.8-17.4)§One was on immunoglobin substitution therapy.7.6 ± 2.9 (1.5-15)‖All were on immunoglobulin substitution therapy.—∗ Defined as diarrhea for at least 3 months with exclusion of infectious cause at the time of inclusion. Based on our recent study,E11Jorgensen S.F. Reims H.M. Frydenlund D. Holm K. Paulsen V. Michelsen A.E. et al.A cross-sectional study of the prevalence of gastrointestinal symptoms and pathology in patients with common variable immunodeficiency.Am J Gastroenterol. 2016; 111: 1467-1475Crossref PubMed Scopus (65) Google Scholar this was not necessarily associated with immune-mediated pathology on biopsy specimens.† Three additional patients with CVID had enteropathy, but they were excluded because of the use of immunosuppressive therapy.‡ Excluding the last month.§ One was on immunoglobin substitution therapy.‖ All were on immunoglobulin substitution therapy. Open table in a new tab Table E2Individual clinical characteristics and special diet for patients with SIgADPatient no.Recurrent infectionsAutoimmunityEnteropathyAsymptomaticSpecial diet11100No21100No31100Vegetarian41100No51000No61000No71000No81000Lactose-free diet91000No101000No111000No120010Lactose-free diet130001No140001Vegetarian Open table in a new tab Table E3Duration of CVID enteropathy and overview of treatment and dietary limitationsPatient no.∗The numbers are randomly allocated and are not corresponding to the CVID numbers used in Table E5.Age at inclusion (y)Duration of enteropathy (y)Dietary limitationsImmunosuppressive drugs14117Gluten-free dietOral budesonide259Since childhood†Exact age at onset not described other than childhood.NoNo3516NoNo44014NoOral budesonide5581NoNo65424NoNo7306NoNo84643Sugar- and lactose-free dietNo9624NoNo10357NoNo11759NoNo12368NoNo13631NoNo145810NoNo∗ The numbers are randomly allocated and are not corresponding to the CVID numbers used in Table E5.† Exact age at onset not described other than childhood. Open table in a new tab Table E4Significantly different taxa between patients with CVID and healthy controls using analysis of composition of microbiomesTaxaPercentile for patients with CVIDPercentile for healthy controls02550751000255075100Firmicutes.Clostridia.Clostridiales.Christensenellaceae.Christensenellaceae R-7 group∗The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in healthy controls compared with patients with CVID.1210857527,98612237652,29116,655Firmicutes.Clostridia.Clostridiales.Lachnospiraceae.Hungatella†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in patients with CVID compared with healthy controls.1423872,9911113125Firmicutes.Clostridia.Clostridiales.Ruminococcaceae.Flavonifractor†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in patients with CVID compared with healthy controls.4611956199,4241614431,985Firmicutes.Negativicutes.Selenomonadales.Veillonellaceae.Veillonella†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in patients with CVID compared with healthy controls.1282201,31427,58514127114,837Proteobacteria.Gammaproteobacteria.Enterobacteriales.Enterobacteriacee.Escherichia-Shigella†The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in patients with CVID compared with healthy controls.153791,63719,8561153712,817Only those taxa that were present in at least 5 samples were included in analysis. For each significantly different taxon, the number of reads per phyla is given in percentiles for healthy controls and patients with CVID. For example, 50th percentile for Flavonifractor means that in half the samples from patients with CVID, 195 or fewer sequences were assigned to Flavonifractor compared with only 14 in healthy controls.∗ The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in healthy controls compared with patients with CVID.† The most prevalent taxa (selecting the taxa that had at least 5 in either disease phenotype at 75 centiles) was increased in patients with CVID compared with healthy controls. Open table in a new tab Table E5Serum IgA levels in the patients with CVIDCVID no.∗The numbers are randomly allocated and are not corresponding to the CVID numbers used in Table E3.Serum IgA1<0.12<0.13<0.14<0.15<0.16<0.17<0.18<0.19<0.110<0.111<0.112<0.113<0.114<0.115<0.116<0.117<0.118<0.119<0.120<0.121<0.122<0.123<0.124<0.125<0.126<0.127<0.128<0.129<0.130<0.131<0.132<0.133<0.134<0.135<0.136<0.1370.1380.3390.3400.3410.3420.3430.5440.5450.5460.6470.6481.1491.4502.0∗ The numbers are randomly allocated and are not corresponding to the CVID numbers used in Table E3. Open table in a new tab Only those taxa that were present in at least 5 samples were included in analysis. For each significantly different taxon, the number of reads per phyla is given in percentiles for healthy controls and patients with CVID. For example, 50th percentile for Flavonifractor means that in half the samples from patients with CVID, 195 or fewer sequences were assigned to Flavonifractor compared with only 14 in healthy controls.
Referência(s)