← Back to Feb 2014

BOS Phenotypes and Classification

links image

Keith C. Meyer, MD, MS
UWHC Lung Transplant and Advanced Pulmonary Disease Program
University of Wisconsin School of Medicine and Public Health
Madison, Wisconsin, USA

When I took on the position of medical director for our University of Wisconsin lung transplant program as it started up in 1988, I faced a steep learning curve. Not only could a host of things go awry up front for lung transplant recipients in the peri-operative period, but the successfully transplanted patient could subsequently develop a plethora of complications that would threaten their allograft function and their post-transplant survival. As the art and science of lung transplantation evolved, it became clear that many patients would have a chronic decline in their lung function and that lung biopsies frequently showed histopathologic changes of obliterative bronchiolitis (OB). Because transbronchial biopsies were often non-diagnostic and surgical lung biopsy would entail significant risk, significant decline in lung function (FEV1) associated with an obstructive physiologic pattern was adopted as a surrogate marker for allograft dysfunction, and this dysfunction was assumed to be due to the presence of OB; hence this syndrome was dubbed bronchiolitis obliterans syndrome (BOS) when FEV1 fell below the threshold of 80% of the best post-transplant FEV1 value, the FEV1 decline was persistent, and the decline in allograft function could not be explained by other factors such as acute cellular rejection or infection [1,2].

In 2008, Ganesh Raghu (medical director of lung transplantation at the University of Washington) and I proposed a project to the American Thoracic Society to establish guidelines for the management of lung transplant recipients. We then sought and received the partnership of the ISHLT and European Respiratory Society (ERS). Paul Corris, Geert Verleden, Jim Egan, Paul Aurora, and Allan Glanville joined the project as co-chairs. We eventually realized that our project was overly ambitious and decided to narrow the focus to the diagnosis and management of BOS. Thirty-seven additional lung transplant experts from transplant centers worldwide accepted our invitation to join our task force committee, and we then thoroughly searched the literature to gather evidence to update our concepts of what BOS truly represents, how it should be diagnosed, and how it can be managed according to the best available evidence. The result is a joint clinical practice guideline (ATS/ISHLT/ERS) that will soon be published in the European Respiratory Journal.

As our project progressed, it became clear that terminology was somewhat of a problem, especially in the light of new information. Do all patients whose FEV1 falls irreversibly below 80% of their best post-transplant FEV1 value have OB as the cause? Is alloimmune rejection the sole cause of BOS? Is BOS the best term to use for delayed (e.g. ≥3 months post-transplant) allograft function decline? Could other factors be recognized that cause or contribute to delayed lung function decline? Articles that were appearing in the published literature often equated BOS with chronic rejection and used the terms BOS, chronic rejection, and chronic lung allograft dysfunction (CLAD) interchangeably.

Chronic lung allograft dysfunction is a term that was occasionally used in the 1980s, but its definition was not clearly established, and it did not make an appearance in the lung transplant literature until 2010. As our task force discussed terminology and new findings, it became clear that calling all cases of delayed allograft dysfunction "BOS" was a gross lumping of all potential causes and cases of delayed allograft dysfunction into a category that assumed that allograft functional decline was solely due to chronic rejection that caused OB. Observations by the Leuven and Munich transplant groups had identified patients whose allograft function declined and met criteria for BOS but could subsequently respond to azithromycin therapy and improve to the point that they regained enough allograft function to re-enter the BOS 0 category; the term "neutrophilic reversible allograft syndrome (NRAD)" was coined because these patients typically had significant neutrophilia in bronchoalveolar lavage (BAL) fluid, and this tended to predict a positive response to azithromycin. Additionally, the Toronto and Leuven groups had identified a substantial subset of "BOS" patients who had a restrictive pattern on pulmonary function testing, tended to have fibrotic infiltrates on thoracic imaging consistent with pleuro-parenchymal fibrosis, and had lung histopathology that showed inflammatory/fibrotic changes with or without evidence of OB; this entity is now perceived to be a form of CLAD that has significant differences in physiology, histopathology, and clinical course in contrast to the "classical" obstructive (reduced FEV1 and FEV1/FVC) BOS entity that was described by previous society statements on BOS, and it has been dubbed "restrictive allograft syndrome (RAS)." It has also become clear that gastroesophageal reflux (GER) with microaspiration can play a significant role in inducing allograft functional decline, and the Duke transplant group has published data showing that recipients who met BOS criteria and had significant GER may improve significantly following anti-reflux surgery. Some of the other potential etiologies have also been identified that include primary graft dysfunction (PGD), autoimmune reactions to glycoproteins expressed by lung cells and tissues (e.g. collagen V), infections (viral, bacterial, fungal), or exposure to high levels of particulate air pollution.

So how should we use new knowledge to classify BOS subsets, and what can we consider to be BOS phenotypes? Perhaps it would be better to adopt the term "CLAD" to replace BOS as the overarching term for delayed/chronic decline in lung allograft function [see reference 3 for our current classification proposal] yet not discard the term BOS, which would be a specific subset of CLAD (note that CLAD should not be used as a synonym for BOS). Subsets (phenotypes) of CLAD could be designated as ("classic") BOS (obstructive physiology without infiltrates on thoracic imaging), azithromycin-responsive allograft dysfunction (ARAD) that is characterized by airway inflammation and improves when neomacrolide therapy is administered, or RAS (restrictive physiology with parenchymal infiltrates). A fourth category could consist of "other causes" and include extra-allograft causes (e.g. diaphragmatic paralysis, pleural disease, or native lung hyperinflation) as well as allograft abnormalities (e.g. anastomotic dysfunction, infection, or primary disease recurrence).

Of course, it must be recognized that these other causes of chronic dysfunction may coexist with BOS, ARAD, or RAS and contribute to allograft functional decline. As an overarching term, CLAD would include all forms of chronic allograft dysfunction following transplant. Additionally, CLAD could also be applied to a transplanted lung that does not achieve a reasonable degree of function following transplantation (e.g. significantly reduced when measured against predicted normal physiologic indices). However, it should be kept in mind that CLAD should be used as a descriptor for a transplanted lung that is losing function or has sustained lack of normal function and not used as a diagnosis. One must also recognize that some patients will not "fit" cleanly into a CLAD phenotype (BOS, ARAD, RAS) and that phenotypes may have significant overlap. Finally, a decline of 10-20% in FEV1 or FVC could be considered to represent "suspected" CLAD as was the case for the 0-p stage in the BOS classification system, and the previously adopted BOS staging system (stages 1, 2, and 3) could continue to be used for CLAD, BOS, and RAS.

One must then ask whether additional phenotypes can be identified. Could early onset BOS with rapid and relentless progression to graft failure be considered to represent a specific phenotype versus late-onset BOS? Can BOS associated with a specific risk factor such as GER, the appearance of a de novo donor-specific antigen, persistent acute rejection, bronchiectasis with persistent infection, a prominent immune response to collagen V, or a community-acquired viral infection be considered to represent specific phenotypes of BOS? Can thoracic high-resolution computed tomography (HRCT) patterns aid or even play a key role in identifying CLAD phenotypes? Can biomarkers be used to refine the identification and differentiation of phenotypes? Sorting these questions out will require an ongoing discussion and ongoing research by the lung transplant community with the goal of being able to identify specific CLAD and BOS phenotypes for which specific therapies of potential benefit can be given to arrest the progression of the disorder and possibly restore seemingly lost function, sustain quality of life, and optimize survival of the lung transplant recipient.

Disclosure Statement: the author has no conflicts of interest to disclose.


  1. Cooper JD, Billingham M, Egan T, et al. A working formulation for the standardization of nomenclature and for clinical staging of chronic dysfunction in lung allografts. International Society for Heart and Lung Transplantation. J Heart Lung Transplant 1993;12:713-716.
  2. Estenne M, Maurer JR, Boehler A, et al. Bronchiolitis obliterans syndrome 2001: an update of the diagnostic criteria. J Heart Lung Transplant 2002;21:297-310.
  3. Verleden GM, Raghu G, Meyer KC, et al. A new classification system for chronic lung allograft dysfunction. J Heart Lung Transplant. 2013 Oct 24. pii: S1053-2498(13)01503-9. doi: 10.1016/j.healun.2013.10.022. [Epub ahead of print]

Share via:

links image    links image    links image    links image