The grading of lung rejection derived initially from animal studies begun in the early 1970s and culminated in a series of elegant works by Prop et al. In these models, it was noted that the insertion of an allograft lung precipitated an inflammatory reaction which was initially localized to the perivascular zones. If uncontrolled, the inflammatory infiltrate progressed to involve all portions of the pulmonary interstitium, including blood vessels and airways, ultimately resulting in a hemorrhagic infarcted lung and extensive vasculitis. With the administration of cyclosporin and/or steroids this inflammatory reaction became muted. When lung transplantation was offered for the treatment of patients with chronic pulmonary disorders, a more regimented classification scheme became necessary for therapeutic decisions. At this point an international group of pathologists convened in Palo Alto, CA to devise a classification scheme which would incorporate the beneficial components of the many schemes in use. This required much cajoling and negotiation, as part of the effort to create a relatively simple classification scheme which could be easily taught, easily reproduced by competent pathologists and could be used for inter-institutional collaborations and comparisons. Ultimately in 1990, this led to the Working Formulation for the Standardization of Nomenclature In The Grading of Lung Rejection, a project sponsored by the International Society for Heart and Lung Transplantation. After five years of use, a second expanded group of pathologists convened in Pittsburgh in 1995 to revise the Working Formulation. The result is the enclosed revised Working Formulation for the Classification of Lung Allograft Rejection. This revision is to be published in the Journal of Heart and Lung Transplantation. It is a simplified system which highlights the degree of airway inflammation as a significant variable in acute lung rejection.

In both the original and revised grading schemes, there are overriding themes. First the alloreactive infiltrate which permeates the transplanted lung affects two major sites - the perivascular zones and the airways, both large and small. It was recognized by all participants that the initial mononuclear infiltrates occur in the perivascular zones, particularly in the perivenular areas, and as their intensity increases, more blood vessels (including arteries) are affected and the infiltrate surrounding these blood vessels becomes more pronounced. At some point the infiltrate expands the adjacent alveolar septae where it is accompanied by reactive change of the alveolar epithelium, and in some instances, cell necrosis with sloughing of these cells and associated proteins into the airspaces. This results in a prominent alveolar macrophage response and hyperreactive pneumocytes. At the same time as the pulmonary vasculature is affected, the airways of the lung become actively infiltrated and inflamed. This affects both large and small airways, and results in epithelial cell necrosis and metaplasia of adjacent preserved epithelium. The dual injury to the vessels and airways provided the impetus for a classification of pulmonary rejection based on injury to these two sites. While these two zones of the lung represent the two major areas of focus, it should be recognized that with persistent high grade rejection, the alveolar septae may also be injured, resulting in areas of interstitial scarring.

Rather than recapitulate the description of the grading scheme indicated in the attached preprint, several key points are presented here. First, as acute rejection progresses in intensity, several alterations occur: (1) increased cellular permeation of the alveolar septae (2) increased numbers of eosinophils (3) increased numbers of neutrophils and alveolar macrophages (4) increased degrees of subendothelial permeation by mononuclear cells and (6) increased degrees of small and large airway inflammation. In general, the vascular and airway alterations occur hand-in-hand and any dissociation of inflammatory infiltrates should raise the possibility of a non-rejection related process, particularly infection. Pulmonary rejection frequently occurs without cardiac rejection in heart-lung recipients (66%) and endomyocardial biopsy is inadequate to evaluate the lung allograft.

In 1995, the pathologists with expertise in lung allograft rejection felt that more emphasis on the airway inflammatory component of acute rejection was necessary. In the past, the subscripts of large and small airway inflammation in acute rejection were largely ignored by clinicians. In the 1995 proposal, it is necessary to comment on the degree of airway inflammation and assign an intensity score to it. This feeling derived from two studies by Yousem et al. First, one study on acute rejection noted that those cases with airway inflammation and intraluminal fibromyxoid tissue were predisposed to develop bronchiolitis obliterans. A second study on lymphocytic bronchitis/bronchiolitis of noninfectious origin suggested an increased risk of obliterative airway disease. In addition to these studies, it is an informal observation that children have much more pronounced airway damage in acute rejection. This may be a consequence of the relatively large amount of bronchial associated lymphoid tissue in the pediatric transplant population and the resulting exuberant mixed lymphocyte reaction that occurs in the wall of these airways.

For those who utilized the old working formulation, lymphocytic bronchitis and bronchiolitis would be diagnosed as "no acute cellular rejection with airway inflammation".

In addition to the prototype rejection reactions described in the Working Formulation, two other forms of acute rejection should be recognized. First, some rejection reactions are rich in eosinophils, to the point where the mononuclear inflammatory infiltrate is obscure. These eosinophilic rejection reactions must be distinguished from infectious causes of parenchymal eosinophilia, primarily fungal infestations. Aspergillus, cocciodiodomycosis, and mycobacteriosis have been described as inducing an eosinophilic pulmonary reaction which can mimic acute rejection. Second, partially treated acute rejection reactions may be characterized by plugs of fibromyxoid connective tissue within the airways and airspaces, resembling an organizing pneumonia. This granulation tissue reaction occurs as a result of alveolar septal and airway epithelial damage. It usually reflects partial resolution of a rejection reaction treated with immunosuppressives. Other cases represent acute rejection superimposed on the proliferative phase of diffuse alveolar damage, resulting from graft ischemic injury. Finally, instances of recurrent native disease occurring in the allograft have been reported, and this can cause diagnostic confusion. Examples of recurrent disease described in the literature include sarcoidosis, giant cell interstitial pneumonia, and lymphangioleiomyomatosis.

The histopathology of bronchiolitis obliterans is described in the subsequent lecture. It is however important to emphasize that acute rejection reactions may be superimposed on underlying bronchiolitis obliterans. It is also worthwhile noting that in patients with bronchiolitis obliterans, airway colonization by bacteria and fungi is common in the late clinical course. This causes much difficulty and consternation to the pathologist who is trying to distinguish chronic inflammation of the airways due to immunologic injury from that induced by persistent pseudomonas or aspergillus colonization of the tracheobronchial tree.

Hyperacute lung rejection, in its classical sense, has not been convincingly reported in the literature e.g. necrotizing vasculitis and hemorrhage soon after insertion of the allograft. If one utilizes a different clinical or pathological definition however, there are reports of cases in which graft failure within days or weeks, has been accompanied by the development of or pre-operative presence of anti-HLA antibodies, anti-endothelial cell antibodies, or high panel reactive antibody measurements. Certainly antibody and complement deposition may play a role in acute rejection in addition to cellular mechanisms - we have just not seen convincing cases where the presumed immune complex deposits have elicited vascular necrosis. Subtle, early cases of hyperacute rejection may also be confused with diffuse alveolar damage from ischemic injury in which neutrophil margination and migration through vessel walls can resemble vasculitis.

In summary, the Working Formulation is not a diagnostic scheme without problems. However it is clinically revelant and does offer a general diagnostic construct for all institutions to report their clinicopathologic data. This discussion has served to clarify certain areas described broadly in the 1990 publication and its 1995 revision.