Prospective Assessment of a Standardized Pathologic Grading System for Acute Rejection in Lung Transplantation

Prospective Assessment of a Standardized Pathologic Grading System for Acute Rejection in Lung Transplantation* Alberto De Hoyos, M.D.; Dean Chamberlain, M.D.; Ricardo Schvartzman, M.D.; Juan Ramirez, M.D.; Steven Kesten, M.D., F.C.C.R; TImothy L. Winton, M.D.; and janet Maurer; M.D.

or

Using the recent standardization of the pathologic de6nitions for acute lung rejectioD, we prospectively evaluated 66 consecutive bronchoalveolar lavages (BAL) and transbronchial biopsy (TBB) specimens in 31 patients after lung transplantation. Clinical indications for bronchoscopies were surveillance (D 44), rejectioD (D 18), and infection (D 4). Bronchoalveolar lavages were obtained from the right middle lobe or lingula in single lung transplant and from both sites in double lung transplant recipients. Cytosmears for differential ceO counts were performed and 400 to 500 cells were counted. Five to eight TBB specimens were taken from two different lobes and stained with hematoxylin-eosin, elastic trichrome, and silver methenamine. Sixty-four of 66 sets of biopsy specimens were satisfactory, but 3 were eliminated because of presence of cytomegalovirus cytopathic changes. Of the remaining 61, rejection was present in 45 (74 percent): grade 1 in !3 (38 percent), grade I in 19 (31 percent), and grade 3 in 3 (5 percent). In 30 of 4! (71 percent) surveillance biopsy specimens, rejection was present, grade I in 18 (43 percent) and grade I or 3 in II (28 percent). In TBBs performed £or clinical suspicion of rejection, IS of 18 TBB specimens (83 percent) showed rejection, grade I in 5 (28 percent) and

grade I or 3 in 10 (55 percent). four biopsies performed for suspicion ofinfection, ODe was oormal and three showed rejection in addition to infection. These three were eliminated from further analysis due to the limitation of the Lung Rejection Study Group criteria in distinguishing rejection from infectioD. Of the 45 episodes of rejection, !4 (53 percent) occurred during the 6rst 3 months posttransplantation, 8 (18 percent) between 3 and 6 months and 13 (29 percent) after 6 months, Percentage of BAL lymphocytosis was signi6cantly elevated in grade I or 3 rejection (28±4) when compared with grade 1 (15±3) or grade 0 (10±3) (p
lung transplantation offers a treatment opI solated tion for advanced pulmonary parenchymal and

characterized, interpretation of biopsy specimens has remained problematic and created nomenclature discrepancies among different institutions. In July 1990, the International Society of Heart Transplantation sponsored a meeting in which pathologists from seven institutions (Lung Rejection Study Group [LRSG]) participated in creating a working formulation for the standardization of nomenclature in the diagnosis of pulmonary rejection. 8 In an effort to assess the usefulness of this standardized pathologic grading system, a prospective evaluation was conducted on a series of transbronchial biopsy (TBB) specimens taken from patients who had undergone lung transplantation in our center. All bronchoscopic specimens were graded according to the LRSG definitions and compared with the results of both bronchoalveolar lavage (BAL)cell analysis and clinical diagnosis.

=

=

=

vascular disease.v" Two-year survival for single and double lung transplantation varies between 58 percent and 69 percent." As improved strategies of graft preservation, surgical techniques, and postoperative immunosuppression are employed, early graft failure and airway complications have become uncommon and infection and rejection are now the most frequent causes of morbidity and mortality" The accurate diagnosis and effective treatment of early graft rejection and infection remain-a significant impediment to extending recipient survival." Acute lung rejection is thought to represent a T-cell-mediated phenomenon in response to major histocompatibility complex antigen differences between donor and recipient." Although the histologic changes of rejection are well *From the Departments of Medicine (Drs. De Hoyos, Schvartzman, Kesten, and Maurer), Surgery (Drs. Ramirez and Winton), and Surgical Pathology (Dr. Chamberlain), The Toronto Hospital, University of Toronto, Toronto, Ontario, Canada. Dr. de Hoyos is supported in part by the Lung Association: Metropolitan Toronto and York Region. Manuscript received June 30; revision accepted October 22.

CMV=cytomegalovirus; DLT=double lung tr~lantation; LRSG = Lung Rejection Study ~roup; SLT=singie lung transplantation; TBB transbronchial biopsy

=

METHODS

lbtients A total of 32 patients prospectively underwent bronchoscopy between October 1990 and March 1991. Sixteen patients had CHEST I 103 I 6 I JUNE, 1993

1813

undergone single lung transplantation (SLT)and 16 had undergone double lung transplantation (DLT). Preoperative diagnosis included interstitial lung disease (9), obstructive lung disease (10), cystic fibrosis (1), bronchiectasis (3), and pulmonary hypertension (3). Criteria for donor and recipient selection and the surgical techniques employed have been described previously 4

control from two lobes of the same lung. Slides were stained with hematoxylin-eosin, elastic trichrome, and silver methenamine. They were examined both for evidence of opportunistic infection and rejection. Transbronchial biopsy specimens and BALs were interpreted blindly by a single investigator (D. C.).

Immunosuppressive Regimen

Histologic condition ofTBB specimens was considered the "gold standard" for diagnosis of rejection. We used TBB specimens to determine the sensitivity and specificity of the clinical diagnosis and of BAL lymphocytosis. Data between patients with and without rejection were calculated using unpaired t test. Paired t test was performed to compare results among patients within the same group. Simple regression analysis was used to determine the correlation between BAL lymphocytosis and rejection on TBB specimens. Values are expressed as mean ± SEM. A p
Immunosuppression was initiated with cyclosporine, 8 to 15 mgj kgjd and decreased progressively by 6 months posttransplant to 5 to 8 mg/kg/d: prednisone, 0.25 to 0.5 mg/kg/d, and azathioprine, 1 to 1.5 mglkgjd, were also given. During the induction phase, methylprednisolone, 500 mg intravenously intraoperatively, and antilymphocyte globulin, 10 to 20 mg/kg/d for 7 days, were also utilized. Rejection episodes rated grade 2 or greater on TBB were treated with augmented systemic corticosteroids (methylprednisolone, 500 to 1,()()() mgjd for 3 days followed by prednisone, 1 mgfkgjd and tapered to the preaugmentation dose over the following 4 weeks). Grade 1 rejection was not treated.

Bronchoscopy Bronchoscopy with BAL and TBB was performed (1) according to a surveillance protocol in asymptomatic patients (every three months during the first year, every 6 months during the second year, and yearly thereafter); (2) whenever symptoms, physical examination, chest radiograph, or pulmonary function studies revealed a significant change in the pulmonary status of the patient suggesting infection and/or rejection (see Table 1); and (3) as followup in patients after treatment of histologically proven grade 2 or greater acute rejection (considered under surveillance in the analysis of the data). Fiberoptic bronchoscopy was performed under topical lidocaine (Xylocaine) anesthesia of the upper aiJway and mild intravenous sedation. The bronchoscope was wedged in a subsegmental bronchus of the right middle lobe and/or lingula and a 2o-ml aliquot of phosphate-buffered saline solution was injected and aspirated through the bronchoscope to remove airway material. One hundred to 150 ml of phosphate-buffered saline solution was then injected in 5O-mlaliquots and aspirated through gentle manual suction. The last 30 to 50 ml of the recovered 8uid was analyzed for opportunistic infection and differential cell count. Polymer (Millipore) filterprepared slides were stained by the Papanicolau method and 400 to 500 cells were counted. Gram and silver methenamine staining was carried out and material was cultured for bacteria, fungus, and virus.

Five to eight TBB specimens were obtained under ftuoroscopic

Table I-Clinical Criteria/or Diagnosing Rejection and OpportuniBtic Infection

Criteria Symptoms Physical signs

Physiology Radiology Laboratory (WBC)

1814

Rejection Dyspnea, fatigue, dry cough Normal or mild fever, crackels, decreased O2 saturation on exercise Normal or decreased FEV. Normal, localized, or diffuse interstitial or alveolar infiltrates Normal or increased

Infection Dyspnea, productive cough Fever, crackels, decreased O2 saturation

Analysis ofData

RESULTS

Sixty-sixbronchoscopies were performed during the study period. Indications were surveillance (n = 44), clinical suspicion of rejection (n = 18), and suspicion of infection (n = 4). Of the 61 bronchoscopies in which satisfactory tissue was obtained, 42 were performed for surveillance of rejection, 18 for suspicion of rejection, and 1 for suspicion of infection. Twelve of the 42 surveillance bronchoscopies were performed as followup after treatment of rejection. Biopsy indications and results are shown in Table 2. Biopsy specimens were deemed satisfactory if they had at least 75 to 100 alveoli, small bronchioles or alveolar ducts, and small blood vessels. Absence of opportunistic infection was also a prerequisite for proper interpretation and grading. Pulmonary Rejection

All 32 patients experienced at least one episode of histologic rejection. There were a total of 48 episodes of rejection. In three of these cases, concomitant cytomegalovirus (CMV) infection precluded appropriate grading and these were eliminated from further analysis. The remaining 45 episodes of histologic rejection were identified among 61 satisfactory biopsy specimens (74 percent); 16 sets of biopsy specimens were normal. Rejection grade 1 was present in 23 (38 percent), grade 2 in 19 (31 percent), and grade 3 in 3 (5percent). Twenty-four of the 45 (53percent) episodes of rejection occurred during the first 3 months posttransplantation, 8 (18 percent) between three and 6 months, and 13 (29 percent) after 6 months. Table 2 Table 2 -

Decreased FEV. Localized or diffuse interstitial or alveolar infiltrates Normal, increased, or decreased

Be" ofTrafllbronchial

Biopliea

Histologic Condition Indication

Normal

Grade 1

Grade 2-3

Surveillance Rejection Infection

12

18 5

12

3 1

10

StandardIzed Pathological Grading System for Rejection in Lung Transplantation (De Hoyos et aI)

60

r •

50 til CD

>.

.

.«: ~ ~ ..

()

40

0

s: a. E >..J

30

et

20

..J

to 0~

1. Photomicrograph shows acute rejection . The finding of lymphocytic aggregates more or less confined to the perivascular zone with no extension into alveolar septae as shown herein indicates a grade 2/4 lesion (hematoxylin-eosin, x 160). FIGURE

."

.. .. J

.: ~ '

.)

10 0

-

f ·.... . I ·.. ·· 0

.. ·..·..··..

f

Grade I

Grade II

Grade 11/

Rejection 3. Percentage of bronchoalveolar lavage (BAL) lymphocytosis in single and double lung transplant recipients. FIGURE

FIGURE 2. Photomicrograph demonstrates acute rejection . The lymphocytic infiltrate is not well defined geographically and alveolar septae are infiltrated, therefore indicating a grade 314 lesion (hematoxylin-eosin, x 160).

depicts the indications for bronchoscopy and the incidence of rejection. Rejection was present in 15 of 18 (83 percent) specimens from TBBs performed for clinical suspicion of rejection, grade 1 in 5 and grade 2 to 3 in 10. Thirty of 42 specimens from surveillance TBBs (71 percent) showed rejection, grade 1 in 18 and grade 2 to 3 in 12. Of those 12, only one was from the group that was being followed after treatment of a recent rejection episode. Thus, 11 of 12 significant rejection episodes in the surveillance group occurred as an incidental pick-up in regularly scheduled postoperative follow-up. Examples of rejection grade 2 and 3 are depicted in Figures 1 and 2. BAL Lymphocytosis

Percentage of BAL lymphocytosis during episodes free of rejection and during rejection episodes is depicted in Figure 3. Bronchoalveolar lavage percent-

age oflymphocytosis was significantly greater in grade 2 to 3 rejection (28 ± 4) than in grade 1 (15 ± 3) and 0 (10 ± 3) (p
In nine patients, complete data were available to compare the change in percentage of BAL lymphocytosis after a reduction in severity of rejection (grade 2 or greater) in response to corticosteroids. Patients typically had another biopsy performed 3 to 4 weeks after the initial procedure . Percentage of BAL lymphocytosis decreased from 29 ± 6 before treatment to 19 ± 3 after treatment; however, this did not reach statistical significance . Clinical Diagnosis

Clinical signs of rejection (Table1)were very reliable for the diagnosis of rejection (specificity, 94 percent); however, 21 episodes of rejection occurred in asymptomatic patients (sensitivity, 46 percent). CHEST I 103 I 6 I JUNE. 1993

1815

Outcome of Patients With Grade 1 Rejection Of the 23 episodes of grade 1 rejection, 5 (22 percent) progressed to grade 2, 4 within 1 month and 1 at 3 months. The initial indication for bronchoscopy in patients who progressed from grade 1 to grade 2 rejection within 1 month was suspicion of rejection. The ind ication in the other patient was surveillance on both occasions . The conditions of 12 patients with 14 episodes of grade 1 rejection have remained stable, while 4 patients have died: 2 of sepsis, 1 from obliterative bronchiolitis, and 1 from an Epstein-Barr virus-related B-celllymphoma. Of the 14 episodes of grade 1 rejection in 12 patients, 8 episodes remained grade 1 rejection and 6 resolved spontaneously to grade 0 on subsequent routine TBB. Opportunistic Infection and Rejection Besides the four patients in whom infection was the indication for bronchoscopy, infection (a positive viral culture or positive bacterial culture in the presence of clinical symptoms) was also present in seven additional patients in whom it was not suspected. Acute rejection was present in six of nine patients in whom CMV was isolated from BAL. Although the diagnosis of rejection was made with confidence based on the presence of perivascular lymphocytic infiltrates, appropriate grading could not be performed in three cases due to the limitations of the LRSG guidelines with respect to the overlap in pathologic changes between infection and rejection. These three cases were eliminated from the analysis. In these three patients, CMV pneumonia was diagnosed on the basis of characteristic intranuclear inclusion bodies (Fig 4). Of interest was an association between CMV and Aspergillus species recovered from BAL. Cytomegalovirus was isolated in 4 of6 patients (66 percent) from whom Aspergillus grew on BAL specimens, but only in 5 of 26 patients without Aspergillus on BAL (20 percent). Three of the four patients with both CMV .. o'.

';

., . "

" .

. . .. g

~

'.. .... .

• •

.., . •

.

.

:~

I

..

" •

'

.,

.

. ..~

-'"

0

..~". .. ... • • •#

I.-,, ' . . . I ,

';,, '

•

.-

..

00

FIGURE 4. Photomicrograph shows inflamed alveoli with alveolar exudates in trans bronchial biopsy material . Note evidence of cytomegalovirus-infected cells (hematoxylin-eosin, X 160).

1816

and Aspergillus have died. One died of invasive aspergillosis, one of CMV, and another of obliterative bronchiolitis. The fourth patient developed a fungus ball in his remaining native lung. A pneumonectomy was performed and the patient remains well 18 months posttransplantation. The other two patients with Aspergillus alone remain alive and well . Complications There were three pneumothoraces after TBBs, one in a SLT recipient and two in DLT recipients. Two cases resolved with tube thoracostomy and one required thoracotomy for persistent bronchopleural fistula. There were no major hemorrhages or deaths. DISCUSSION

Before 1988, the diagnosis of acute pulmonary rejection was confirmed late and retrospectively when a clinical deterioration associated with a decrease in pulmonary function and/or radiographic changes resolved after a course of augmented immunosuppression. In the last few years, however, several studies Table 3-Working Formulation for Claai./ication and Grading of Pulmonary Rejection Acute Rejection Grade O-No significant abnormality Grade 1- Minimal acute rejection With evidence of bronchiolar inflammation Without evidence of bronchiolar inflammation With large airway inflammation No bronchioles present Grade 2- Mild acute rejection With evidence of bronchiolar inflammation Witbout evidence of bronchiolar inflammation With large airway inflammation No bronchioles to evaluate Grade 3-Moderate acute rejection With evidence of bronchiolar inflammation Without evidence of bronchiolar inflammation With large airway inflammation No bronchioles to evaluate Grade 4-Severe acute rejection With evidence of bronchiolar inflammation Without evidence of bronchiolar inflammation With large airway inflammation No bronchioles to evaluate Active Airway Damage Without Scarring Lymphocytic bronchitis Lymphocytic bronchiolitis Chronic Airway Rejection Bronchiolitis obliterans-subtotal Active Inactive Bronchiolitis obliterans - total Active Inactive Chronic Vascular Rejection Vasculitis Formal LRSG classification of lung rejection . Detailed histology of various grades of acute rejection outlined in text. From Reference 8.

Standardized Pa1hoIoglcalGrading System lor Rejeclion In Lung Transplantation (De Hoyos st 8/)

have documented the value and safety of TBB specimens taken via Bberoptic bronchoscopy as an objective means for repeatedly obtaining tissue for assessment of rejection and response to therapy.9.IO To facilitate communication among centers performing lung transplantation and to better understand the characteristics and behavior of lung rejection, the LRSG established a standardized grading system for the histologic diagnosis of pulmonary rejection." To our knowledge, this study represents the first report of a prospective assessment of this grading system. Minimal rejection (grade 1) is characterized by very occasional small perivascular lymphocytic aggregates. Mild rejection (grade 2) is reflected by more numerous perivascular aggregates that usually are larger in size than those of minimal rejection. There may be an associated lymphocytic bronchiolitis. In moderate rejection (grade 3), both vessels and bronchioles are involved but, in addition, the lymphocytic infiltrate extends away from these structures into alveolar septae. These changes are even more marked in severe rejection (grade 4) where, in addition, alveolar damage and alveolar exudates appear (Table 3). It is evident from this classification that grading of acute rejection is closely linked to quantitative tissue lymphocytosis. If this is reflected in BAL results, the BAL might be useful in diagnosing rejection or assessing response to treatment of rejection. Using the LRSG grading system, we found a significant increase in BAL lymphocytosis during episodes of grade 2 and 3 acute rejection. A BAL lymphocyte differential of at least 20 percent was found to be highly specific but not sensitive for the diagnosis of rejection. The sensitivity and specificity of using clinical criteria alone was quite similar to that found for a 20 percent BAL lymphocytosis. Thus, it does not appear that BAL alone will be useful in the diagnosis of acute rejection, but it may have more potential application in treatment follow-up. In follow-up BALs in our data (n = 9), we were able to show a decrease in percent lymphocytosis following augmented immunosuppression, but the decrease was not statistically significant with the number of patients we were able to study More research in this area is necessary A limitation of the present study is that we had only three cases of grade 3 rejection and none of grade 4 rejection. There are two possible explanations for this. One is that these degrees of rejection occur uncommonly; a second is that close follow-up of the patients results in early bronchoscopy and treatment of rejection. If grade 3 and 4 rejection reflect progression of untreated rejection (although this has not been established), treating early mild and moderate rejection might result in fewer cases of severe pathologic findings. In all but one of the treated episodes of grade 2 and 3 rejection we documented, there was a reduc-

tion in the severity of the rejection. The exception, a DLT recipient, had persistent grade 2 and 3 rejection despite augmented immunosuppression and succumbed to obliterative bronchiolitis. We arbitrarily elected not to treat grade 1 rejection. However, we did closely follow these patients and have recorded the outcome of documented grade 1 rejection events. Fourteen of the 23 patients with episodes of grade 1 rejection (61 percent) remained in stable condition; 5 (22 percent) progressed to grade 2 rejection and 4 patients died: 3 without known progression of rejection and 1 of obliterative bronchiolitis. Since obliterative bronchiolitis is thought to be a form of chronic rejection, this case may represent progression of untreated rejection. Of interest is that in four of the five patients who progressed to grade 2 rejection, the initial indication for bronchoscopy was suspicion of rejection, suggesting that they were symptomatic and may have required treatment even with minimal pathologic changes. This might represent sampling error in that larger or more numerous specimens may have shown higher degrees of rejection. Alternatively; grade 1 changes that are likely to progress may be accompanied by clinical symptoms. Further study is needed to better characterize the implication of grade 1 histologic features. Potential problems with the interpretation of TBB specimens despite a standardized pathologic grading system warrant consideration. Transbronchial biopsy specimens are small and may not be representative of a process occurring in the lung, although the relatively large number of biopsy specimens now recommended may overcome this limitation. Minimal, mild, and moderate rejection are angiocentric and peribronchiolar in nature and inadequate sampling (absence of vessels and bronchioles) might therefore preclude accurate interpretation. Different severities of rejection could conceivably occur in different areas of the lung, thus making the interpretation dependent on the area sampled by the forceps. Finally, diagnosis and grading of rejection may be difficult or not possible in the presence of infection. Isolation of CMV from BAL alone, however, does not preclude the diagnosis and grading of pulmonary rejection. In our study, most episodes of rejection occurred during the initial three months posttransplantation. However, 29 percent of rejection episodes occurred after the sixth month, when most patients were clinically asymptomatic. This study confirms reports that histologically significant but clinically unsuspected rejection may occur late posttransplantation. II As noted above, another Bnding in our study was the simultaneous presence of rejection and opportunistic infection in tissue and BAL. This is of practical significance since intervention requires opposite approaches. In addition, three of four patients with both CHEST I 103 I 6 I JUNE, 1993

1817

CMV and Aspergillus isolated from BAL have died, supporting previous reports of the poor outcome of this association." Furthermore, the profile of BAL cell differential in cases of opportunistic infection (CM\; Pneumocystis carinii) and rejection is similar with predominance of lymphocytosis. 13 Despite the potential and actual limitations, we have found the working formulation proposed by the LRSG a helpful means to evaluate TBB specimens in lung transplant recipients. Bronchoalveolar lavage lymphocytosis appears to correlate with the presence and severity of rejection, but in itself does not appear sufficient to diagnose or characterize rejection episodes. Acute lung rejection occurs more frequently than clinically suspected early and late after transplantation. Several questions remain unanswered and require further study Of particular interest is the issue of minimal rejection (grade 1) that occurs often asymptomatically. Whether such cases should be treated, closely followed, or ignored remains an unanswered question and requires further study The role of the BAL cell differential is limited and definite diagnosis of rejection requires TBBs. New directions such as cell markers and functional analysis of immune cells may prove more rewarding for diagnosis of acute rejection. REFERENCES

1 Grossman RF, Frost A, Zamel N. Patterson GA, Cooper JD, Myron PR, et a1. Results of single-lung transplantation for bilateral pulmonary fibrosis. N Engl J Med 1990; 322:727-33

1818

2 Emery ~ Graif JL, Hale K, Eales F, Von Rueden TJ, Pritzker MR, et a1. Treatment of end-stage chronic obstructive pulmonary disease with double lung transplantation. Chest 1991; 99:533-37 3 Levine SM, Gibbons WJ, Bryan CL, Walling AD, Brown ~ Bailey SR, et a1. Single lung transplantation for primary pulmonary hypertension. Chest 1990; 98:1107-15 4 De Hoyos A, Patterson GA, Ramirez JC, Winton TL, Miller J, Maurer J. Lung transplantation: early and late results. J Thorac Cardiovasc Surg 1992; 103:295-306 5 Ramirez JC, De Hoyos A, Patterson GA. Airway complications after lung transplantation. Clin Cardiovasc Surg (in press) 6 Griffith B~ Hardesty RL, Trento A, Paradis IL, Duquesnoy RJ, Zeevi A, et a1. Heart-lung transplantation: lessons learned and future hopes. Ann Thorac Surg 1987; 43:6-16 7 Krensky AM, Weiss A, Crabtree G, Davis MM, Parham ~ T lymphocyte-antigen interactions in transplant rejection. N Engl J Moo 1990; 322:510-17 8 Yousem SA, Berry G, Brunt EM, Chamberlain D, Hruban RH, Sibley RK, et al. A working formulation for the standardization of nomenclature in the diagnosis of pulmonary rejection: Lung Rejection Study Group. J Heart Transplant 1990; 6:593-600 9 Higenbottam T, Stewart S, Penketh A, Wallwork J. Transbronchial lung biopsy for the diagnosis of rejection in heart-lung transplant recipients. Transplantation 1988; 46:532-39 10 Starnes VA, Theodore J, Oyer PE, Stinson EB, Moreno-Cabral CE, Sibley R, et a1. Pulmonary infiltrates after heart-lung transplantation: evaluation by serial transbronchial biopsies. J Thorac Cardiovasc Surg 1989; 98:945-50 11 Clelland C, Higenbottam T, Otulana B, Stewart S, Igboaka G, Scott J, et a1. Histologic prognostic indicators for the lung allografts of heart-lung transplants. J Heart Transplant 1990;

9:177-86

12 Rubin RH. Impact of cytomegalovirus infection on organ transplant recipients. Rev Infect Dis 1990; 12(suppl 7):8754S766 13 Dauber JH, Paradis IL, Dummer JS. Infectious complications in pulmonary allograft recipients. Coo Chest Med 1990; 11:291308

StandardIzed Pathological GradingSystem tor Rejection in Lung li'anspIantation (De Hoyosst 8/)