INTRODUCTION

TOP ABSTRACT INTRODUCTION CASE REPORT DISCUSSION REFERENCES

Hyperacute rejection has been well documented with kidney (1, 2), heart (3, 4), and liver (5) transplants in both experimental systems and patients. The morphologic changes resulting from hyperacute rejection have been described for renal and heart transplants (2, 3). Within minutes to hours of vascular anastomosis, the transplanted organ becomes grossly edematous, mottled, and cyanotic. Platelet thrombi form within capillaries and small arterioles followed by neutrophil margination and fibrin thrombi. These events are followed by obvious endothelial damage, interstitial edema, hemorrhage, and infarction.

The exact mechanism of hyperacute rejection is not clear, but it appears to involve preformed antibodies against major allograft antigens, usually HLA or ABO. The initial targets for the antibodies appear to be the endothelial cells. Intuitively, these are the first allograft cells accessible to the recipient's circulating antibodies. One model proposes that complement and neutrophil-mediated damage of the endothelial cell and exposure of the basement membrane initiate the accumulation of vasoactive substances, sequestration of blood elements, and finally activation of the coagulation, fibrinolytic, and kinin-forming systems (2). More recent models propose a more active role for the endothelial cells (6, 7). Upon antibody binding, endothelial cells release numerous chemotactic and coagulation factors. The death of the endothelial cells occurs later. Eventually, these processes result in the disruption of small blood vessels, leading to ischemic necrosis and extravasation of fluid and blood into the interstitium.

Lung transplantation has been used for the treatment of various diffuse pulmonary diseases. Advances in surgical technique and immunosuppression have increased the 1-yr survival rate after single- or double-lung transplantation to greater than 70% (St. Louis International Lung Transplant Registry, April 1996 Report). Common early complications include graft failure, acute cellular rejection, and infection. Experimental hyperacute rejection of xenograft pulmonary allografts has been described. In one model, an isolated pig lung is mechanically ventilated and perfused with human blood, leading to rapid graft failure within 3 h associated with pulmonary edema, hemorrhage, and fibrin/platelet thrombi (8). The immediate clinical findings and the acute (3 d post-transplant) pathologic findings have been reported for the first case of hyperacute rejection of a pulmonary allograft (9). This second case report confirms the immediate clinical findings and provides first documentation of the immediate (within 4 h) pathologic findings in the hyperacute rejection of the lung.

	CASE REPORT

TOP ABSTRACT INTRODUCTION CASE REPORT DISCUSSION REFERENCES

Clinical Summary

The patient was a 50-yr-old woman with severe COPD since 37 yr of age without known etiology except for a 25 pack-year history of smoking, which she quit in her early 30s. A pretransplant evaluation for panel reactive antibodies (PRA) was negative. The patient underwent an orthotopic single right lung transplant without complication with an ischemic time of 2 h 56 min. The usual lung preservation and single lung transplantation methods were used as described by Trulock (10). Post-transplant evaluation of patient serum and donor lymphocytes showed cross-matching compatibility. The patient received standard immunosuppressive regiments intravenously consisting of azathioprine (2 mg/kg/d), methylprednisolone (500 mg intraoperative bolus followed by 0.5 mg/kg/d), and cyclosporine (titered to blood level of 250 to 350 ng/ml). Her post-operative course was complicated by biopsy-proven acute cellular rejection for which she received methylprednisolone (1 mg/ kg/d for 3 d) and anti-lymphocytic globulin (ALG). Subsequently, the patient developed diffuse alveolar damage and progressive graft failure. Eleven days after the first transplant, the patient underwent a second orthotopic single right lung transplant with an ischemic time of 3 h 14 min. Immediately after closure, pink frothy fluid was noted in the right orifice of double-lumen endobronchial tube. The pink fluid accumulated rapidly and reached as much as 100 ml every 5 to 10 min. The left bronchus remained clear. The patient was reopened to reveal a boggy and edematous transplanted lung. A wedge biopsy was obtained. Pulmonary vein pressure before and after the anastomosis was equal. Within a period of 60 min, the patient's blood pressure, oxygen saturation decreased rapidly while her potassium level changed from 4.9 to 11.2, with EKG showing wide complex QRS. The patient had multiple ventricular arrhythmias; despite a prolonged resuscitation, she was declared dead 4 h after anastomoses of the second lung transplant.

The donor left lung was transplanted without complication at another institution. That patient expired 7 d post-transplantation with disseminated cytalomegalovirus (CMV) infection. The two donors and the recipient patient for this case report were all CMV seropositive without serologic evidence of recent infections.

Pathology

Sections of the first allograft stained with hematoxylin-eosin (H&E) showed hyaline membranes and organizing fibrinous material within alveolar spaces consistent with organizing diffuse alveolar damage. There was mild to moderate perivascular lymphocytic infiltrate consistent with acute rejection. No fibrin/platelet thrombus or neutrophilia was present.

The wedge biopsy of the second allograft taken 4 h after anastomoses and reperfusion revealed marked interstitial neutrophilic infiltrate with platelet and fibrin thrombi in small arterioles (Figures 1a and 1b). Martius scarlet blue and trichrome stains confirmed the findings of platelet/fibrin thrombi. The alveolar spaces were clear with no evidence of edema or neutrophils. The bronchioles were normal with no evidence of peribronchial edema or inflammation. No perivascular edema or lymphocytic infiltrate was present. No evidence of CMV infection was present by morphology or by paraffin immunohistochemistry using antibodies against CMV antigens (DAKO mouse monoclonal; DAKO Ltd., Carpenteria, CA).

View larger version (138K): [in this window] [in a new window]   Figure 1.   Second pulmonary allograft, wedge biopsy, 4 h postanastomoses. (a) Platelet thrombus with interstitial neutrophilia (hematoxylin-eosin stain; original magnification: ×400). (b) Fibrin and platelet thrombus and interstitial neutrophilia (hematoxylin-eosin stain; original magnification: ×200).

A full autopsy was performed 6 h postmortem. The second allograft weighed 800 g with dark red coloration and firm consistency. The cut surface showed patchy poorly defined areas of hemorrhagic consolidation. All anastomoses sites were intact and widely patent. In contrast, the native left lung weighed 560 g with pale pink color and a hypercrepitant consistency. The cut surface showed emphysematous changes and a 4-cm wedge-shaped peripheral hemorrhagic infarct.

H&E sections of the second transplanted lung showed platelet/fibrin thrombi similar to those seen in the wedge biopsy. In focal areas, there was interstitial neutrophilia without evidence of alveolar involvement. Other areas showed varying amounts of alveolar edema and neutrophilic infiltrate (Figures 2a and 2b). The peribronchial region was essentially free of inflammation with no evidence of bronchopneumonia. There was no evidence of hyaline membrane formation, necrosis, peribronchial edema, perivascular edema, lymphocytic infiltrate, or CMV infection. In contrast, H&E sections of the native lung showed emphysematous change, congestion, and interstitial edema (Figure 2c). Fibrin/platelet thrombi and interstitial neutrophilia were not observed.

View larger version (74K): [in this window] [in a new window]   Figure 2.   Autopsy, 6 h postmortem. (a) Second pulmonary allograft with interstial neutrophilia and sparse intra-alveolar neutrophilic infiltrate (hematoxylin-eosin stain; original magnification: ×200). (b) second pulmonary allograft with increased intra-alveolar neutrophils and edema (hematoxylin-eosin stain; original magnification: ×200). (c) native lung with no specific change (hematoxylin-eosin stain; original magnification: ×400).

Paraffin immunohistochemistry performed on the second transplanted lung using antibodies against immunoglobulin heavy chains showed deposition of antibodies on the endothelial surface and within the vasculature walls. No evidence of CMV infection was present using antibodies against CMV antigens.

The hilar and peribronchial nodes showed partial effacement of the nodal architecture by a lymphocytic infiltrate with scattered transformed lymphocytes, immunoblasts, and plasma cells. In situ hybridization for a latent Epstein-Barr viral transcript (EBER) showed scattered positivity (data not shown). These findings were considered sufficient for a diagnosis of early post-transplant lymphoproliferative disorder (PTLD) and categorized as polymorphous B-cell hyperplasia.

Immunology

The HLA types of the patient and donors were determined to be (1) patient: A1,2; B8,35,w6; DR3,6; DQ1,DQ2; (2) first donor: A3,23(9); B57(17),w4; DR7,11(5); DQ2,DQ3; and (3) second donor: A3,32; B7,44(12),w4,6; DR2,11(5); DQ1,DQ3.

The patient's serum was negative in cytotoxicity testing against a panel of lymphocyte donors (0% panel reactive antibody-PRA) prior to the first transplant and was cross-match compatible with the first allograft. Treatment of the patient with ALG prevented a microcytotoxicity cross-match analysis with postengraftment sera. To overcome this problem the patient's serum subsequent to the second transplant was analyzed by HLA flow cytometry. This analysis revealed IgG antibodies that reacted strongly to both the second donor's B and T lymphocytes, greater than four and ten times background respectively (Figure 3). The two allografts shared four potentially immunologic HLA antigens: A3, Bw4, DR11(5), and DQ3. Patient's sera were tested against a panel of more than 20 different B and T lymphocytes by flow cytometry. Although many of the panel members were positive (some members reacting 20-fold over background), no single anti-HLA reactivity could be established in the sera. However, the possibility of an anti-Bw4 could not be excluded.

View larger version (23K): [in this window] [in a new window]   Figure 3.   HLA flow cytometry for reactivity to second donor's lymphocytes.

	DISCUSSION

TOP ABSTRACT INTRODUCTION CASE REPORT DISCUSSION REFERENCES

The clinical and pathological findings in this case are consistent with hyperacute rejection of the lung. Rapid (minutes to hours) failure of the graft, an edematous and cyanotic appearance, fibrin/platelet thrombi within the small arterioles and capillaries of the allograft, and pronounced neutrophilia are constant early findings in hyperacute rejection of other organs. These findings are specific to the second right allograft and are not seen in the first allograft or in the remaining native lung, ruling out some unusual systemic disorder. The rapid onset of these findings in only the second allograft, the recipient's positive CMV serology prior to any transplantation, and the absence of pathologic evidence of CMV infection argue strongly against an acute CMV infection.

Evidence for other known causes of rapid graft failure are not present. Both intraoperative and postmortem examination revealed intact and patent anastomoses. Ischemic reperfusion injury (IRI) of the lung is unlikely since rapid graft failure was not seen with the left side donor allograft transplanted at a different institution. Furthermore, no report of abundant pink frothy fluid in the bronchus has been linked to IRI. The histologic findings of IRI (11), which consists of peribronchial and perivascular edema with small number of lymphocytes were not present. Small amounts of neutrophilic infiltrate can be seen in IRI, but usually after 24 h.

Varying amounts of pulmonary edema with alveolar neutrophilic infiltrate were present during postmortem examination but not in the immediate postmortem wedge biopsy. The pulmonary edema is probably not a postmortem artifact given the intraoperative boggy appearance of the allograft and the copious production of pink, frothy fluid. The timing and the quality of the fluid production appear identical to that reported by Frost and associates (9). Although unlikely, the possibility of alveolar neutrophils being a postmortem change cannot be completely excluded. However, in our experience, alveolar neutrophils are not seen as a postmortem change even in failed lung transplant cases.

The prominent neutrophilic infiltrate seen in this case is discrepant from that seen in the ex vivo xenograft experimental model (8). Perfusion of human whole blood through a pig lung resulted in rapid graft failure and fibrin/platelet thrombi without leukocyte infiltrate. The absence of neutrophilic infiltrate most likely reflect dysfunction of the human neutrophils during the long term storage; whole human blood was stored as long as 30 d, which greatly exceed the typical 8-h half-life of neutrophils in vivo. Although very unlikely, a species difference between the pulmonary allografts cannot be completely excluded as an explanation for the discrepancy.

The geographic discordance of the pulmonary edema and alveolar neutrophils indicates an evolution in the patterns of hyperacute rejection. The relationship among the variable histologic patterns, different rates of damage, and different rates of pulmonary perfusion is not clear. By analogy with the findings in renal hyperacute rejection, the sequence of events might include (1) immediate endothelial lifting off the basement membrane, (2) platelet thrombi, neutrophilia, and edema, (3) fibrin thrombi, (4) alveolar neutrophils, (5) parenchymal hemorrhage and infarction. Features 1 to 4 may occur over a time course of minutes to hours depending on the level and reactivity of pathogenic antibodies. Parenchymal hemorrhage and infarction are absent in this case, and in the previously reported case, probably reflecting an inherent difference between lung allograft and other organ allografts. Alternatively, the hemorrhage and infarction may occur and resolve between the time points of the two cases (6 h verses 3 d), although this seems unlikely. It appears that hyperacute rejection in pulmonary allografts progress to diffuse alveolar damage (9).

In hyperacute rejection, the antibodies are usually against HLA or ABO antigens. In this case, the obvious risk factor for sensitization is the prior lung transplant. We have demonstrated by HLA flow cytometry that new antibodies were formed between the engraftment of the first and the second transplants. Despite extensive efforts, we were unable to confirm the target antigen.

This patient also had evidence of early PTLD. Compared with other organ transplants, lung transplants appear to have an increased risk for PTLD (12). EBV is a known stimulator of B-cell proliferation. Whether or not viral infection with EBV can increase B cell reactivity and alloantibody production and thereby increase the frequency or severity of graft rejection is unknown. It is interesting to speculate that the early PTLD in this patient may have been another factor contributing to the production of an alloantibody.

In summary, this case documents the second occurrence of hyperacute rejection in a pulmonary allograft. Pulmonary edema and copious production of frothy, pink fluid from the bronchial orifice of the allograft appear to be an early and constant clinical finding. The pathologic features of hyperacute rejection except for the hemorrhage and necrosis are similar in the lung allografts and other vascular allografts. Future advances in graft preservation and in cross-match methods may allow prospective HLA matching and may prevent this complication.