Bronchiolitis obliterans syndrome, hypogammaglobulinemia, and infectious complications of lung transplantation

Transcription

1 Bronchiolitis obliterans syndrome, hypogammaglobulinemia, and infectious complications of lung transplantation Daniel C. Chambers, MBBS, MRCP, FRACP, MD, a,b Belinda Davies, BN, b Ann Mathews, BSc, a Stephanie T. Yerkovich, BSc(Hons) PhD, a,b and Peter M. Hopkins, MBBS, FRACP a,b From the a School of Medicine, The University of Queensland; and the b Queensland Lung Transplant Service, The Prince Charles Hospital, Brisbane, Queensland, Australia. KEY WORDS: bronchiolitis obliterans syndrome; hypogammaglobulinemia; fungal infection; community-acquired respiratory viral infection; lung transplantation BACKGROUND: Because infection has been associated with the development of bronchiolitis obliterans syndrome (BOS), we hypothesized that post-transplant hypogammaglobulinemia would be associated with infection and BOS. METHODS: Cross-sectional levels of serum immunoglobulins were measured on 2 occasions in our transplant cohort and models developed to explain serum immunoglobulin levels and BOS-free survival. RESULTS: A total of 139 patients (median age, 46.6 years) were evaluated at 47 months (range, months) after transplant, and 87 were re-evaluated at 72 months ( months). Of this cohort, 44% were immunoglobulin (Ig) G deficient and levels remained stable across the study period, and 27% were IgA deficient and levels fell slightly over time (p ¼ 0.003). Both immunoglobulin classes were lower in patients with a history of invasive fungal infection, whereas IgA levels were lower in patients with a history of community-acquired respiratory viral infection. Low IgG was independently associated with shorter BOS-free survival (hazard ratio, 0.79; 95% confidence interval, ; p o 0.001). CONCLUSION: Serum immunoglobulin deficiency is common after lung transplantation and is associated with community-acquired respiratory viral infection, invasive fungal infection, and BOS. J Heart Lung Transplant 2013;32:36 43 r 2013 International Society for Heart and Lung Transplantation. All rights reserved. Lung transplantation is the only treatment option available for many patients with end-stage pulmonary disease, but long-term survival is compromised by the development of bronchiolitis obliterans syndrome (BOS). Although the strongest risk factor for the development of BOS remains a history of acute rejection, 1 infectious complications may also contribute. Cytomegaloviral and community-acquired respiratory viral infections (CARV) 2,3 as well as colonization with Aspergillus spp 4 and Pseudomonas aeruginosa 5 have all been implicated in BOS onset and progression. Mucosal surfaces cover an area of more than 400 m 2 and harbor a vast population of potential pathogens 5 that are Reprint requests: Daniel C. Chambers, MBBS, MRCP, FRACP, MD, Queensland Lung Transplant Service, The Prince Charles Hospital, Brisbane, QLD 4032, Australia. Telephone: Fax: address: daniel_chambers@health.qld.gov.au /$ - see front matter r 2013 International Society for Heart and Lung Transplantation. All rights reserved. kept in check by the barrier function of these surfaces, secreted bactericidal peptides and proteins, and by the humoral arm of the immune system, in particular immunoglobulins (Ig) A and IgG. As the only commonly transplanted organ with a mucosal surface, the lung allograft may be particularly susceptible to the impaired mucosal defense associated with immunoglobulin deficiency. Hypogammaglobulinemia is common after solid-organ transplantation and has been variously implicated as a causal factor in the infectious complications associated with these procedures. After lung transplantation, hypogammaglobulinemia (defined as IgG levels below age-specific ranges for pediatric patients 6 or o 6 g/liter 7 or 7 g/liter 8,9 for adult patients) is almost the norm, with a prevalence of between 48% and 73%. 6 9 Severe IgG deficiency (o 4 g/liter) is less common, with a prevalence between 8% and 15%, but has been associated with a higher risk of pneumonia, bacteriemia,

2 Chambers et al. BOS, Hypogammaglobulinemia, and Infection 37 and invasive fungal infection. 8,9 Given the prevalence of hypogammaglobulinemia after solid-organ transplantation and the putative role of infection in BOS pathogenesis, we hypothesized that IgG and IgA deficiency would be common after lung transplantation and would be associated not only with infectious complications but also with BOS. Methods The study was approved by The Prince Charles Hospital Human Research and Ethics Committee. Patients and study design This study is a mixed retrospective review of pre-transplant immunoglobulin levels and a prospective cross-sectional assessment of post-transplant immunoglobulin levels. We assessed serum immunoglobulin levels in 139 lung transplant patients at our center in 2009 (97% of the entire cohort) and reassessed 87 patients (63%) 2.5 years later. All patients were censored in November In a sub-set of 75 patients, antibody levels were measured at the pre-transplant assessment. The study excluded patients receiving supplemental immunoglobulins. Immunoglobulin measurements Total IgA, IgG, and IgM were measured by nephelometry (Siemens Healthcare Diagnostics, Deerfield, IL), with normal reference range of 47 g/liter for IgG, 41 g/liter for IgA, and g/liter for IgM. Immunoglobulin deficiency was defined as a serum level below the lower limit of the predicted normal range. Clinical information Clinical parameters were collected including demographic data, immunosuppression, history of basiliximab or rituximab use, rejection burden (sum of A grades), history of proven (by polymerase chain reaction of nasopharyngeal swab/aspirate or bronchoalveolar lavage for respiratory syncytial virus, parainfluenza, influenza, human metapneumovirus, or adenovirus) CARV, cytomegalovirus infection, P aeruginosa colonization status (defined as isolation of P aeruginosa on at least 2 bronchoalveolar lavage samples taken at least 1 month apart), history of definite or probable invasive fungal infection 10 and BOS status. 11 Immunosuppression protocols, consisting of a calcineurin inhibitor, mycophenolate mofetil, and prednisolone, have been uniform since the inception of the program in Induction therapy is not routinely used. Basiliximab is reserved for patients requiring a delay in calcineurin inhibitor exposure. Rituximab has been used rarely in patients with hyperacute or acute humoral rejection. Anti-microbial prophylaxis consists of targeted antibacterial drugs in the first 7 to 10 days after transplant, anti-fungal prophylaxis (voriconazole or itraconazole) for the first 3 months, and at times of augmented immunosuppression thereafter. Statistical analysis Results are presented as median and interquartile range (IQR). Statistical analysis was performed using STATA 11 software (StataCorp, College Station, TX) and p o 0.05 was considered statistically significant. Group differences were assessed by Kruskal-Wallis test, Mann-Whitney U test, Pearson chi square, or the Fisher exact test, as appropriate. Paired samples were compared by the Wilcoxon signed-rank test. Correlations between variables were determined using Spearman s r. Simple and multivariate linear regressions were performed with IgG or IgA as the dependent variable. Simple linear regression analysis was initially used to evaluate the relationship between variables and immunoglobulin levels, and those variables with p o 0.1 were then subjected to multiple linear regression analysis. The time from transplant to BOS was modeled using Cox proportional hazard regression. Potential predictors included demographic factors, IgG, IgA, and IgM measured post-operatively, and previous infection, considered as a time-dependent covariate. The final model was obtained by forwards and backwards selection, retaining covariates where inclusion or exclusion changed the coefficients of other predictors by 4 10% or where predictors were statistically significant at a ¼ Results Full demographic data for the cohort are reported in Supplementary Table 1 (available on the JHLTonline.org Web site). Patients (48% women) were a median age of 46.6 years (IQR, years) at transplantation. The diagnosis was cystic fibrosis in 34% and chronic obstructive pulmonary disease (COPD) in 37%. The transplant procedures were bilateral in 86%, single-lung in 7%, heart-lung in 5%, and heart-lung-liver in 2%. At the census, 51% of the cohort had a diagnosis of BOS grade Z 1. Immunoglobulin levels fall after transplantation and remain low At the first measurement at a median of 47 months after transplant, 44% of the patients were IgG-deficient (7.4 [IQR, ] g/liter) and 27% were IgA-deficient (1.4 [IQR, ] g/liter; Figure 1). IgG and IgA levels were only modestly correlated (r ¼ 0.349, p o 0.001). Only 2 patients (identical twins with a history of cystic fibrosis and without associated IgG deficiency) were totally IgAdeficient (o 0.07 g/liter). IgM levels were better maintained after transplant, with a median IgM level of 0.9 g/liter (IQR, g/liter) and with only 13% of the cohort deficient. Levels of IgG and IgA were below normal in 20 patients (14%), and all 3 immunoglobulin classes were deficient in 6 (4%). Paired immunoglobulin levels before and after transplant were available for 47% (IgG), 52% (IgA), and 54% (IgM) of the cohort (Figure 2 and Supplementary Figure 1, available on the JHLTonline.org Web site). Age, sex, diagnosis, and transplant type were not different between those with and without pre-transplant immunoglobulin levels (data not shown). Pre-transplant, median (% deficient) IgG, IgA and IgM levels were 11.4 g/liter (9%), 2.5 g/liter (6%), and 1.1 g/liter (5%), respectively. The peri-transplant fall in IgG, IgA, and IgM levels was highly significant (p o for each).

3 38 The Journal of Heart and Lung Transplantation, Vol 32, No 1, January 2013 demographic factors or immunoglobulin levels (as assessed at the first measurement) between those who did and did not have repeated immunoglobulin levels measured (data not shown). Matched analysis showed stable IgG levels (p ¼ 0.958) after transplant, whereas IgA (p ¼ 0.003) and IgM (p ¼ 0.004) fell slightly during this time (Figure 2 and Supplementary Figure 1, available on the JHLTonline.org Web site). All subsequent analyses were performed using the first measurement (n ¼ 139). Figure 1 Serum immunoglobulin (Ig) G, IgA, and IgM levels are low after transplantation. Total IgG (left y axis), IgA (right y axis), and IgM (right y axis) levels were measured in 139 lung transplant patients. Median values are shown by the thick grey line, the lower limit of the normal range is shown by a black line, and the percentage of the cohort deficient in each immunoglobulin subtype is indicated at the bottom of the graph. To understand the effect of time after transplant on immunoglobulin levels, serum immunoglobulin levels were reassessed at a median of 31.4 months (IQR, months) after the first measurement. Repeat levels were available in 87 patients (63%); 32 (23%) had died and 20 (14%) were lost to follow-up. There was no difference in the Demographic predictors of post-transplant hypogammaglobulinemia IgG (r ¼ 0.264, p ¼ 0.002) and IgM (r ¼ 0.181, p ¼ 0.033) levels declined with age, but this age-related decline was not seen with IgA (Supplementary Figure 2, available on the JHLTonline.org Web site). There was no sex-related difference in IgG or IgA levels, but median IgM levels were lower in male recipients (0.8 g/liter vs 1.2 g/liter, p ¼ 0.002, Supplementary Figure 3, available on the JHLTonline.org Web site). IgG (p o 0.001) and IgA (p ¼ 0.04) levels were both lower in those with a pre-transplant diagnosis of COPD (Supplementary Figure 4, available on the JHLTonline.org Web site). IgA, but not IgG or IgM, levels were significantly positively associated with mycophenolate mofetil dose at Figure 2 Serum immunoglobulin (Ig) G and IgA levels fall after transplant, and although IgG remains stable thereafter, serum IgA falls with time after transplant. Paired pre- and post-measurement 1 transplant IgG (n ¼ 65) and IgA (n ¼ 72) levels are shown with a solid line denoting the median. Similarly, IgG and IgA levels are plotted for each measurement time post-transplant in those who had repeated measurements (n ¼ 87). Matched data were compared using Wilcoxon signed rank test, with significance indicated.

4 Chambers et al. BOS, Hypogammaglobulinemia, and Infection 39 Figure 3 Serum immunoglobulin (Ig) G and IgA levels are lower in patients with bronchiolitis obliterans syndrome (BOS). Levels for serum (Left) IgG and (Right) IgA are plotted for patients with stable BOS 0 (n ¼ 63), patients who developed BOS (New BOS) during the study period (n ¼ 30), and patients with BOS Z 1(n¼43), with median values and significance indicated. the time of antibody measurement (r ¼ 0.269, p ¼ 0.003, Supplementary Figure 5A, available on the JHLTonline.org Web site), and patients receiving tacrolimus had lower median IgA levels (1.3 g/liter) than patients receiving cyclosporine (1.8 g/liter, p ¼ 0.002, Supplementary Figure 5B, available on the JHLTonline.org Web site); however, no correlation was observed between the serum levels of these immunosuppressive agents and IgA levels. We did not observe an association between choice or dose of calcineurin inhibitor and serum IgG or IgM. Serum IgG, IgA, and IgM levels were not associated with transplant type, use of basiliximab or rituximab (used in 2 patients), rejection burden, or prednisolone dose (data not shown). Hypogammaglobulinemia is associated with the prevalence and incidence of BOS To investigate the effect of hypogammaglobulinemia on subsequent graft outcome, patients were categorized according to their BOS status between 2009 and censoring. During this period, 30 patients (22%) developed new onset BOS, 63 (45%) remained BOS 0, 43 (31%) remained BOS Z 1, and 3 died before their status could be determined. IgG levels were lower in those patients with BOS in 2009 (6.7 [IQR, ] g/liter, p o 0.001) and in those who subsequently developed BOS (6.4 [IQR, ] g/liter, p ¼ 0.008) compared with patients who remained BOS 0 (8.6 [IQR, ] g/liter, Figure 3). IgA levels were also lower in patients with BOS (1.3 [IQR, ] g/liter) than in those without BOS in 2009 (1.6 [IQR, ] g/liter, p ¼ 0.020; Figure 3); however, low IgA levels were not associated with the subsequent development of BOS. In those patients where a second immunoglobulin measurement was available, IgG levels did not change over time for any group (p , data not shown); however, IgA fell significantly in the group that remained BOS 0 (p ¼ 0.003), but not in those with established or new-onset BOS (p , data not shown). We did not observe an association between IgM levels and BOS (data not shown). The multivariate analysis is summarized in Tables 1 and 2. The independent predictors of low post-transplant serum IgG were a pre-transplant diagnosis of COPD and BOS (Table 1). Independent predictors of low post-transplant serum IgA were receiving a lower daily dose of mycophenolate and having a history of CARV (Table 2). Predictors for shorter BOS-free survival are summarized in Table 3. Univariate predictors were increasing age at transplant, a single-lung transplant, a pre-transplant diagnosis of COPD, and having a more significant rejection history. In contrast, higher serum IgG and IgA levels were associated with longer BOS-free survival. In a multivariate model, having a single-lung transplant (hazard ratio, 2.8; 95% Table 1 Factors Significantly Associated With Serum Immunoglobulin G Level After Transplant Variables a b (95% CI) p-value Univariate analysis Age at antibody measurement ( to 0.010) Pre-transplant diagnosis of COPD 1.91 ( 2.86 to 0.96) o0.001 Presence of BOS 1.64 ( 2.57 to 0.72) History of invasive fungal infection 1.20 ( 2.20 to 0.18) Multivariate analysis Pre-transplant diagnosis of COPD 1.47 ( 2.42 to 0.51) Presence of BOS 1.31 ( 2.23 to 0.39) BOS, bronchiolitis obliterans syndrome; CI, confidence interval; COPD, chronic obstructive pulmonary disease. a Only variables with p o 0.01 are included in the Table; r 2 of the multivariate linear regression ¼

5 40 The Journal of Heart and Lung Transplantation, Vol 32, No 1, January 2013 Table 2 Factors Significantly Associated With Serum Immunoglobulin A Level After Transplant Variable a b (95% CI) p-value Univariate analysis History of CARV infection 0.44 ( 0.86 to 0.03) Daily MMF dose 0.33 (0.09 to 0.57) Use of tacrolimus 0.66 ( 1.15 to 0.17) History of invasive fungal infection 0.54 ( 0.98 to 0.09) Multivariate analysis Daily MMF dose 0.29 (0.046 to 0.546) History of CARV infection b 0.37 ( 0.74 to 0.006) CARV, community-acquired respiratory viral; CI, confidence interval; MMF, mycophenolate mofetil. a Only variables with p o 0.01 are included in the Table; r 2 of the multivariate linear regression ¼ b CARV and invasive fungal infection were strongly inter-related, but because CARV generally preceded fungal infection, CARV was retained in the model. confidence interval, ; p ¼ 0.013) was associated with reduced BOS-free survival, whereas having higher serum IgG levels was associated with longer BOS-free survival (hazard ratio, 0.8; 95% confidence interval, ; p o 0.001). Serum IgA was not an independent predictor of BOS-free survival. There was a trend for greater rejection burden to be associated with shorter BOS-free survival. Hypogammaglobulinemia is associated with an increased incidence of viral and fungal infection To determine if the association between BOS and hypogammaglobulinemia is mediated by a greater burden of infection, we determined the incidence of CARV, cytomegaloviral infection, invasive fungal infection, and P aeruginosa colonization. There were only 2 episodes of cytomegaloviral infection, 48% of the cohort had experienced at least 1 episode of CARV, and these individuals had lower serum IgA levels (p ¼ 0.009, Figure 4A). Notably the 2 individuals in our cohort who are totally IgA deficient have experienced far more (total of 12) episodes of proven CARV than any other patient. In contrast, IgG levels were not different in patients with a history of CARV (Figure 4A). Similarly, immunoglobulin levels were not different in patients colonized with P aeruginosa (45% of the cohort, Figure 4B). Finally, the 32% of patients who experienced at least 1 episode of invasive fungal infection had significantly lower IgG (p ¼ 0.012) and IgA (p ¼ 0.019) levels than those who had not (Figure 4C). Discussion We have found that serum IgG and IgA levels both fall after transplantation, with many patients having levels below the reference range, and that immunoglobulin deficiency places the lung transplant recipient at risk of CARV infection, invasive fungal infection, and BOS. Our findings suggest that immunoglobulin deficiency should be excluded in any patient presenting with 1 or more of these post-transplant complications. Several groups have identified a similarly high prevalence (48% to 73%) of hypogammaglobulinemia after lung transplantation. 6,8 10 Presumably post-transplant hypogammaglobulinemia relates to the commencement of immunosuppression; however, with only 1 exception, where lower IgG levels were associated with mycophenolate use, 10 no studies have been able to demonstrate a correlation between the degree of immunosuppression and IgG levels. 6,8,9 Nevertheless, inadvertent effects of T cell-targeted immunosuppressive therapy on B cells remain the most likely cause for impairment of humoral immune defense. Mycophenolic acid has profound cytostatic effects on B and T cells, whereas corticosteroid and calcineurin inhibitor treatment impair the activated T-cell synthesis of interleukin 2 required by B cells to differentiate into immunoglobulin-producing plasma Table 3 Factors Significantly Associated With Time to Bronchiolitis Obliterans Syndrome Variable HR (95% CI) p-value Univariate analysis Age at transplant 1.01 ( ) Single lung transplant 2.82 ( ) Pre-transplant diagnosis of COPD 1.74 ( ) Rejection burden (sum of A grades) 1.14 ( ) Post-transplant IgG level 0.80 ( ) o0.001 Post-transplant IgA level 0.79 ( ) Multivariate analysis Single lung transplant 2.73 ( ) Post-transplant IgG level 0.79 ( ) o0.001 Rejection burden (sum of A grades) 1.17 ( ) CI, confidence interval; COPD, chronic obstructive pulmonary disease; HR, hazard ratio; Ig, immunoglobulin.

6 Chambers et al. BOS, Hypogammaglobulinemia, and Infection 41 Figure 4 Serum immunoglobulin (Ig) G and IgA levels are lower in patients with community-acquired respiratory viral (CARV) and invasive fungal infection. Levels of serum (Left) IgG and (Right) IgA were assessed in patients with and without a history of (A) CARV, (B) Pseudomonas aeruginosa colonization, and (C) invasive fungal infection. Patients with a history of CARV have lower IgA, whereas patients with a history of invasive fungal infection have lower IgG and IgA. cells. Given this anti B-cell activity, the observed positive relationship between serum IgA levels and mycophenolate dose is likely related to the effects of time after transplant on both, rather than a true association. Only 1 study has assessed serum IgA status in the setting of lung transplantation. Robertson et al 9 found that 4 of 31 pediatric lung transplant recipients (12.2%) had low serum IgA levels and that this was a significant risk factor for invasive aspergillosis. The best evidence suggests that lung mucosal IgA is largely polymeric and derived from local plasma cells, with a more limited contribution by circulating, monomeric IgA derived from bone marrow plasma cells. 12,13 It is therefore perhaps surprising that we identified an independent effect of circulating IgA on risk of CARV. Our results imply that the measurement of circulating IgA is a surrogate for similar deficiency of locally produced IgA or that the role of circulating IgA in lung mucosal defence is greater than previously thought.

7 42 The Journal of Heart and Lung Transplantation, Vol 32, No 1, January 2013 The demographic factors predicting lower IgG levels were closely aligned with those previously reported. It is well known that IgG falls with age and that low IgG is more common in patients with COPD 6,10 ; however, our study found that serum IgA is also lower in patients with COPD. IgM is also known to be lower in male recipients 14 ; however, we found no sex differences with respect to serum IgG levels, in contrast to Kawut et al, 8 who identified female sex as a risk factor for severe hypogammaglobulinemia. We found that hypogammaglobulinemia and low IgA were more common in patients who had experienced an episode of invasive fungal infection or CARV, respectively. Two previous studies have identified hypogammaglobulinemia as a risk factor for bacterial, fungal, and cytomegaloviral infection, 6,7 but no studies have examined the risk of CARV in patients with low immunoglobulin levels. Although patients with low immunoglobulin levels are more likely to experience both BOS and these infectious complications, we cannot determine whether the increased infectious burden is related primarily to immunoglobulin deficiency or to the increased incidence of infection in BOS. 15,16 However, because low serum IgA was not independently associated with BOS, we can conclude from our data that low IgA appears to be a risk factor for CARV infection. Univariate analysis showed IgA and IgG deficiency were both associated with shorter BOS-free survival, but we found that low serum IgG was the strongest independent predictor for shorter BOS-free survival and that patients with subsequent new-onset BOS were more likely to have low IgG, but not IgA, levels. Taken together, our data suggest that older patients who receive allografts for COPD may be at a particularly high risk of developing hypogammaglobulinemia and to experience shortened BOS-free survival, particularly if they have received a single-lung transplant. How low serum IgG leads to an increased risk of BOS is not clear, but did not appear related to colonization with P aeruginosa. 17 We postulate that hypogammaglobulinemia leaves the allograft open to colonization or infection with other organisms (eg, Aspergillus spp 4 ), which then adversely affect allograft function in the longer-term. One previous study, by Goldfarb et al, 7 examined the effect of immunoglobulin deficiency on long-term outcomes and demonstrated worse survival in patients with low IgG. However, it was not clear whether this difference was statistically significant, or whether it was related to BOS status. A number of limitations to this study need to be highlighted. We assessed immunoglobulin levels in a cross-sectional fashion and, therefore, cannot draw conclusions regarding the causality of the significant associations identified. However, our finding that patients with the subsequent development of BOS had lower baseline IgG levels provides some support for the idea that hypogammaglobulinemia is a risk factor for BOS development. We have, as far as is possible given the study design, demonstrated that IgG levels are stable over time, so that the proportional hazard assumption is not violated. For IgA, because levels fall slightly over time, the multivariate modeling will, if anything, have underestimated an effect. As with any cohort study, biases can be introduced if there are significant dropouts, era effects, or a systematic data loss. Owing to the careful follow-up of our cohort and the universal treatment protocols used since the inception of our program (all patients are followed up at our center with no shared-care), we believe most biases have been avoided. Because we have a very low threshold for performing polymerase chain reaction on respiratory secretions in suspected CARV and computed tomography imaging in patients with symptoms suggestive of invasive fungal infection, case detection bias is also likely to be minimized. Despite more than 2 decades of research, the longer-term survival of patients with lung allografts has not improved, with most of the late deaths related to the development of BOS. Our findings provide a new avenue of investigation for research groups focusing on BOS pathogenesis. Of particular interest for our group are the mechanisms underlying the association between hypogammaglobulinemia and BOS, particularly given the growing body of evidence linking B-cell function to longterm allograft outcomes. 18 The role of immunoglobulin replacement therapy for patients with BOS and hypogammaglobulinemia is currently unknown. Owing to the nature of this therapeutic product (immunoglobulins are isolated from pooled plasma that has been donated by 1,000 to 100,000 people), any recommendation for IgG replacement in deficient patients would require an evidence level that could only be generated by a suitably powered randomized controlled trial. Although the risk of transmission of infectious agents has been minimized, it is still real, particularly in the case of parvovirus B Of equal concern are the high cost and the risk of cardiovascular and renal complications, particularly in patients who have poor baseline renal function. 19 In conclusion, immunoglobulin deficiency is very common after lung transplantation, particularly in older patients with a pre-transplant diagnosis of COPD. IgG deficiency should be suspected in patients with infectious complications and BOS, whereas IgA deficiency should be excluded in patients with CARV. Disclosure statement The authors thank Jonathan Bleier, Biostatistician from Queensland Institute of Medical Research, for statistical advice and support. Some of this work was presented at the Thirtieth Anniversary Meeting and Scientific Sessions of the International Society for Heart and Lung Transplantation, Chicago, Illinois, April 20 24, None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the presented manuscript or other conflicts of interest to disclose. Supplementary data Supplementary data are available in the online version of this article at JHLTonline.org. References 1. Glanville AR, Aboyoun CL, Havryk A, Plit M, Rainer S, Malouf MA. Severity of lymphocytic bronchiolitis predicts long-term outcome after lung transplantation. Am J Respir Crit Care Med 2008;177: Gottlieb J, Schulz TF, Welte T, et al. Community-acquired respiratory viral infections in lung transplant recipients: a single season cohort study. Transplantation 2009;87:

8 Chambers et al. BOS, Hypogammaglobulinemia, and Infection Khalifah AP, Hachem RR, Chakinala MM, et al. Respiratory viral infections are a distinct risk for bronchiolitis obliterans syndrome and death. Am J Respir Crit Care Med 2004;170: Weigt SS, Elashoff RM, Huang C, et al. Aspergillus colonization of the lung allograft is a risk factor for bronchiolitis obliterans syndrome. Am J Transplant 2009;9: Backhed F, Ley RE, Sonnenburg JL, Peterson DA, Gordon JI. Hostbacterial mutualism in the human intestine. Science 2005;307: Robertson J, Elidemir O, Saz EU, et al. Hypogammaglobulinemia: Incidence, risk factors, and outcomes following pediatric lung transplantation. Pediatr Transplant 2009;13: Goldfarb NS, Avery RK, Goormastic M, et al. Hypogammaglobulinemia in lung transplant recipients. Transplantation 2001;71: Kawut SM, Shah L, Wilt JS, et al. Risk factors and outcomes of hypogammaglobulinemia after lung transplantation. Transplantation 2005;79: Yip NH, Lederer DJ, Kawut SM, et al. Immunoglobulin G levels before and after lung transplantation. Am J Respir Crit Care Med 2006;173: Husain S, Mooney ML, Danziger-Isakov L, et al. A 2010 working formulation for the standardization of definitions of infections in cardiothoracic transplant recipients. J Heart Lung Transplant 2011;30: Estenne M, Maurer JR, Boehler A, et al. Bronchiolitis obliterans syndrome 2001: an update of the diagnostic criteria. J Heart Lung Transplant 2002;21: Twigg HL 3rd. Humoral immune defense (antibodies): recent advances. Proc Am Thorac Soc 2005;2: Pilette C, Ouadrhiri Y, Godding V, Vaerman JP, Sibille Y. Lung mucosal immunity: immunoglobulin-a revisited. Eur Respir J Crisp HC, Quinn JM. Quantitative immunoglobulins in adulthood. Allergy Asthma Proc 2009;30: Billings JL, Hertz MI, Wendt CH. Community respiratory virus infections following lung transplantation. Transpl Infect Dis 2001;3: Gavalda J, Len O, San Juan R, et al. Risk factors for invasive aspergillosis in solid-organ transplant recipients: a case-control study. Clin Infect Dis 2005;41: Botha P, Archer L, Anderson RL, et al. Pseudomonas aeruginosa colonization of the allograft after lung transplantation and the risk of bronchiolitis obliterans syndrome. Transplantation 2008;85: Kirk AD, Turgeon NA, Iwakoshi NN. B cells and transplantation tolerance. Nat Rev Nephrol 2010;6: Looney RJ, Huggins J. Use of intravenous immunoglobulin G (IVIG). Best Pract Res Clin Haematol 2006;19:3-25.