Can we decloak how infections drive complications after lung transplantation?

Infections are a major risk factor for chronic lung allograft dysfunction (CLAD), which is the main attributable long-term cause of mortality in lung transplant (LTx) recipients.^1–3 Gram-negative organisms such as Pseudomonas aeruginosa are particularly notorious. In experimental models of lung transplantation, Pseudomonas was shown to facilitate a G-CSF-mediated neutrophilia, followed by effector T cell responses, thus contributing to acute rejection.⁴ Additionally, Pseudomonas also increases humoral risk factors for CLAD.⁵ In many patients, Pseudomonas, once colonized, is hard to eradicate.^6,7 The mechanisms by which Pseudomonas evades the host immune response,⁸ thus driving CLAD, remain to be completely understood. Understanding these mechanisms would allow tailored therapies to ideally eradicate Pseudomonas, and if not, at least modulate the host immune response to mitigate both the short- and long-term deleterious consequences of infection.

With this goal in mind, Divithotewala and Pham et al assessed serum ‘cloaking’ antibodies (cAb) against P.aeruginosa in a single-center, retrospective cohort analysis of 123 LTx recipients.⁹ These antibodies target the lipopolysaccharide (LPS) O-antigen of the gram-negative bacterial cell wall, and increase susceptibility of the host to infection by blocking the classical pathway of complement-mediated lysis.¹⁰ They do this presumably by completely covering (“cloaking”) the bacterial surface, thus creating a physical barrier that prevents insertion of the membrane attack complex.¹¹ Such an effect has been observed in other infections such as Neisseria, Vibrio, Salmonella and Klebsiella. Effects are, at least in part, attributed to the ‘Neisser-Wechsberg’ phenomenon wherein specific factors against these bacteria contribute to this blocking effect, and can be adsorbed out of the sera using homologous strains.¹¹ In patients with bronchiectasis harboring P.aeruginosa — this effect, known as ‘antibody-dependent enhancement (ADE)’ — has been attributed to excess production of O-antigen-specific IgG2 antibodies and is associated with worse respiratory function.¹⁰ It was thereby logical when the authors had previously reported that salvage plasmapheresis in a LTx recipient with refractory P.aeruginosa infection was associated with a reduction in cAb titres, restoration of in vitro serum-mediated killing of the infecting P. aeruginosa strain, and clinical improvement.¹² This observation formed the basis of their current hypothesis that cAb against P.aeruginosa would be associated with worse CLAD-free survival post-LTx.

Utilizing 424 biobanked serum samples from 123 LTx recipients among whom 72.4% (89/123) were colonized with P. aeruginosa following transplantation, Divithotewala and Pham et al detected cAb in 40.7% (n = 50/123) of these recipients. Eighty-four percent of patients with serum cAb (n = 42/50) had detectable antibodies within the first year post-LTx and the median time to initial cAb detection post-LTx was 109 days. The most prevalent subclasses of antibodies were the presence of both IgG2 and IgA in 40.0% (n = 20/50) of patients, and 90% of the recipients (18/20) had cAb detected prior to CLAD onset. Serum cAb was independently associated with an increased risk of CLAD, and separately, with an increased risk of all-cause mortality. Many of these cAb were present at the time of the first draw, thus, limiting their ability to discern the time to development of these antibodies, and the temporal relationship between infection and this immune response. Nevertheless, this relationship held true not only in recipients with a primary diagnosis of cystic fibrosis (CF), but also in the non-CF cohort, suggesting the long-term complications of infection on graft survival are independent of the underlying disease, a finding concordant with our prior report.⁵

Patients with CF are a unique population as they are frequently colonized with Pseudomonas as part of the natural history of their underlying disease,^13,14 posing the possibility of the development of cAb prior to lung transplantation. The authors point out that the cAb were present in many of the first sample collections, the impact of the development of cAb prior to transplant is unknown and indistinguishable from the development of de novo cAb post-transplant. The study also found a diagnosis of CF and younger age were both associated with the development of cAb but was unable to disentangle the general association of CF diagnosis and younger age to determine if these are independent risk factors. Patients with CF have been shown to have improved survival following lung transplantation over patients with other diagnoses,¹⁵ yet the authors found in their study that the presence of cAb is associated with the development of CLAD and all-cause mortality. Therefore, it begs the question if outcomes for patients with CF who undergo lung transplantation would be further improved in the absence of cAb. In this study, however, there were insufficient sample sizes to distinguish a difference in the development of CLAD and all-cause mortality between patients with CF who developed cAb and those who did not.

The authors defined colonization as the presence of P.aeruginosa from respiratory samples, including both bronchoalveolar lavage (BAL) and sputum specimens, on two or more occasions at least 3 weeks apart following lung transplantation. In the total cohort, 72% of patients were considered to have Pseudomonas colonization with no statistical difference in colonization rates between the groups with or without cAb. As Pseudomonas colonization was similar regardless of the presence of cAb, it remains unclear the specific trigger for the development of cAb. Whether chronic colonization with Pseudomonas, the specific strain of Pseudomonas including the lipopolysaccharide characteristics which can alter the expression of the O-antigen structure (“smooth” vs. “rough”),¹⁶ the severity of Pseudomonas infection and subsequent tissue damage, or another unrelated factor leads to the development of cAb remains to be elucidated.

A question arises then, whether these cloaking antibodies are potentially an indirect surrogate marker of a broader alloimmune response. This is especially relevant given that the O-specific antigen polysaccharides elicit strong antibody responses. We have previously shown that lung transplant recipients with a history of Pseudomonas aeruginosa are at a higher risk of developing donor-specific antibodies (DSA), and such antibodies (especially against MHC Class II antigens) portend worse CLAD-free survival.⁵ As the authors do not routinely measure DSA post-lung transplantation at their center, their study was unable to assess if the risk conferred by these cloaking antibodies is related to or independent of DSA. Additionally, their study was not designed to address when these antibodies develop, and whether their persistence (and thereby, the adverse long-term outcomes) is dependent on chronic colonization. This is relevant because although these antibodies block complement-mediated killing,^10,11 LPS in itself is a known inducer of complement activation, which has been associated with chronic rejection post-lung transplantation.^17,18 Hence, whether CLAD is a result of these antibodies helping P.aeruginosa evade complement-mediated killing and facilitating their persistence, versus P.aeruginosa in itself promoting complement-mediated injury, either directly or indirectly (e.g., via inducing other antibodies that can activate complement), remains to be answered.

This study by Divithotewala and Pham et al is thus a key observation in improving our understanding of how infections are associated with worse long-term outcomes after lung transplantation and will hopefully spur further prospective observational, mechanistic, and therapeutic studies to mitigate these poor outcomes. Specifically, it will be necessary to dissect the temporal nature of the development of these cloaking antibodies. For example, are they present prior to transplantation in those colonized with Pseudomonas? Or do these develop after transplantation? Is their development concurrent with the development of autoimmune and alloimmune responses that are known to be associated with rejection?¹⁹ And importantly, will eradication of the antecedent stimuli (e.g., Pseudomonas in itself) or these antibodies improve CLAD-free survival, or are these simply epiphenomena, and thus, should we be targeting CLAD independently of these cloaking antibodies? We look forward to “decloaking” the answers to these questions in the years to come, and thus, breaking the vicious cycle between infection and CLAD.