Brief Communications Pulmonary Vascular Permeability after Lung Transplantation A Positron Emission Tomographic Study1-3
JAMES D. KAPLAN, ELBERT P. TRULOCK, JOEL D. COOPER, and DANIEL P. SCHUSTER4
During the past 10yr, human lung transplantation has rapidly become a valuable therapeutic option for end-stage lung disease. Among many important issues, the so-called "reimplantation response" and allograft rejection stand out as two processes of particular importance. Although both are characterized by histologic evidence of lung injury, each process has a distinctive clinical course (1). The reimplantation response has been defined as the "morphologic, roentgenographic, and functional changes that occur in a lung transplant in the early postoperative period as a result of surgicaltrauma, ischemia, denervation, lymphatic interruption, and other injurious processes (exclusiveof rejection)" (2). This response is usually diagnosed by exclusion, i.e., when basal or perihilar infiltrates occur in the early postoperative period and cannot be attributed to rejection, infection, mucus plugging, or cardiogenic pulmonary edema. The severity of the response usually reaches a maximum by Day 4, and clears rapidly thereafter (3). The diagnosis is often confirmed retrospectively when the observed abnormalities resolve without specific therapy. In contrast, allograft rejection rarely occurs during these first few days after transplantation. After the first week, rejection is suspected when fever, dyspnea, malaise, leukocytosis, hypoxia, or new infiltrates occur, and the diagnosis is confirmed clinically by a prompt response to intravenously administered corticosteroids and/or histologically by a perivascular lymphocytic infiltrate in transbronchiallung biopsies (4). Interestingly, surveillance transbronchial biopsy studies after lung transplantation indicate that unsuspected allograft rejection can frequently be present without clinical signs and symptoms (5). Despite these distinctive features, these two complications of transplantation have been remarkably difficult to evaluate clinically using currently available noninvasive methods (1). We have developed a means to noninvasively evaluate pulmonary vascular permeability using the quantitative nuclear medicine imaging technique of positron emission tomography (PET) (6). This technique has been used to evaluate lung injury as a result of the adult respiratory distress syndrome (7), lobar pneumonia (8), chronic cigarette smoke exposure (9), and interstitial lung disease (10). Because 954
SUMMARY Weevaluated pulmonary vascular permeability In 15patients after lung transplantation (21 allografts) by measuring the pUlmonary transcapiilary escape rate (PTCER) for Ga·68·labeled transferrin, using positron emission tomography. Seven recipients (four unilateral, three bilateral lung transplants) were studied within 3 days of transplantation, and each developed hypoxemia and allograft Infiltrates consistent with the "relmplantatlon response." PTCER was higher In subjects studied within 1 day than In those studied at a later time, and fell In seven allografts studied serially. The initial PTCER also correlated (r = 0.77) with length of Ischemic (preservation) time, even In the three subjects with bilateral allografts. Eight other recipients (five unilateral, three bilateral transplants) were evaluated for possible organ rejection at least 1 wk after transplantation. PTCER was normal In patients without clinical or histologic evidence of rejection, and It was elevated In recipients with rejection. PTCER fell each time after treatment for rejection with Increased Immunosuppression In the three patients studied serially. These data suggest that positron emission tomography measurements of PTCER might be a useful way to evaluate both the relmplantation response and organ rejection after lung transplantation. AM REV RESPIR DIS 1992; 145:954-957
both the reimplantation response and allograft rejection are characterized by histologic and radiographic evidence oflung injury, weused PET to determine whether changes in pulmonary vascular permeability could be detected during either condition and whether these changes correlated with successful treatment. Twenty-threePET scans wereperformed on 15adult single- and double-lung transplant recipients. Age, sex, indication for transplantation, and other clinical parameters are indicated in tables 1 and 2. The protocol for this study was approved by the Washington University Human Studies Committee. To evaluate the reimplantation response (defined operationally for this study as hypoxemia requiring supplemental oxygen and new radiographic infiltrates in the allograft), PET scans were performed within 72 h of surgery on seventransplant recipients who were stable for transportation to the scanner. Three of the seven recipients had bilateral transplants, so that a total of 10 lung allografts were studied (Reimplantation Group). Serial scans were repeated in five recipients (seven of 10 allografts) 4 to 8 days later. Eleven allografts in eight other recipients were studied at least 1wk after transplantation to evaluate pulmonary vascular permeability during rejection (Rejection Group). PET scans were performed within the 24 h before subjects underwent bronchoscopy and transbronchial biopsy, either for routine rejection surveillance or for a clinical suspicion of rejection. Histologic evidence of a perivascular mononuclear cell infiltrate with negative microbiologic cultures was considered diagnostic for rejection. Serial studies were performed in three patients after treatment for rejection with intravenously administered methylprednisolone.
PET scan data were acquired on the SUPERPETT I scanner (11),and displayed as seven transaxial tomographic slices with a center-to-center separation of 15mm and an intrinsic in-plane resolution of 18 mm. The methods and theory of PET measurements have been reviewed recently (3). Each PET study included a background scan, a transmission scan, and a series of emission scans after the injection of radioisotope. Sternal landmarks were used to guide patient positioning in the scanner, and these same landmarks were used in the case of serial studies in the same patient. The background and transmission scans wereperformed to correct subsequent emission scans for errors introduced by tissue attenuation during the tissue activity measurement. Because photon attenuation is proportional to tissue density, the transmission
(Received in original form August 2, 1991 and in revised form September 27, 1991) 1 From the Respiratory and Critical Care Division, Department of Internal Medicine, Washington University School of Medicine, St. Louis, Missouri. 2 Supported by Grants R01 HL-41943 and POI HL-13851 from the National Institutes of Health and by Grant DE-FG02-87ER60512 from the Department of Energy. 3 Correspondence and requests for reprints should be addressed to Daniel P. Schuster, M.D., Washington University School of Medicine, Respiratory and Critical Care Division, 660 S. Euclid, Box 8052, St. Louis, MO 63110. 4 Established Investigator of the American Heart Association and a Career Investigator of the American Lung Association.
955
BRIEF COMMUNICATION
TABLE 1 REIMPLANTATION RESPONSE STUDIES Patient No. 1 2 3 4 5 5 6 6 7 7
(yr)
Sex
Side
Indication for Transplant
Ischemic Time
Hours from Reperfusion to Scan
Hours to Follow-up Scan
61 22 53 35 21 21 42 42 53 53
M F F F F F M M F F
L L L R L R L R L R
COPD Bleo Fibrosis COPD EG CF CF AAT AAT Bronchiectasis Bronchiectasis
3' 32" 3' 50" 5' 20" 3' 30" 7'17" 5' 20" 6' 0" 4' 16" 4' 35" 7' 9"
53 70 10 41 24 24 34 34 27 27
238 251 108
Age
118 118 124 124
Definition of abbreviations: L = left; R = right; COPO = chronic obstructive pulmonary disease; Bleo sinophilic granuloma; CF = cystic fibrosis; AAT = alpha-1-antitrypsin deficiency.
= bleomycin;
EG
= Eo-
TABLE 2 REJECTION STUDIES Patient No. 1 2 3 4 5 6 7 8
Age (yr)
Sex
Side
Indication
47 57 46 53 35 37 28 35
F F F F F F F F
R L R L B R B B
COPD COPD COPD COPD COPD PPH IPF AAT
Definition of abbreviations: POD = postoperative day; B pulmonary fibrosis. For other definitions, see table 1.
Rejection Suspected
POD for Initial Scan
Follow-up Performed
+ +
37 7 17 58 30 190 358 104
+
+ + +
= both; PPH = preliminary pulmonary
scan also provides a density image, resembling a standard X-ray-computed tomographic image, which can be used to define anatomic regions of interest (12). Radioisotope was prepared according to previously described methods (12). As much as 6 mCi of 68Ga-citrate were injected intravenously. After injection, 68Ga-citrate rapidly dissociates and the 68Ga then avidly binds to circulating transferrin, serving both as a plasma marker initially and subsequently as a tracer for the measurement of protein flux into the extravascular space. Serial emission scans were collected for 60 min beginning 2 min after injection. Time-activity curves were measured for the PET activity in the lung regions of interest and within the cardiac blood pool on each emission scan and decay-corrected to the time of tracer injection. The blood pool activity was analyzed by two sequential linear regressions to determine the best fit as either a monoexponential or biexponential expression for the blood time-activity curve (12). The tissue and blood time-activity curves were then analyzed according to a two-compartment mathematical model (8, 12). Using this model, a vascular-to-extravascular rate constant (K1) was calculated. Because protein flux is affected by surface area available for exchange, K 1 was normalized to regional blood volume, which was assumed to be proportional to surface area. Regional blood volume was calculated from the initial 2-min (68Ga]transferrin scan, when virtually all tracer activity was still intravascular. Regional pulmonary transcapillary escape rate (PTCER) was then expressed as K1/blood volume. Lung density was calculated by scaling the attenuation scan to the known density of blood, so that the cardiac blood pool had a density of 1.05 g/ml (13). This resulted in images of regional total lung density (TLD), which included both vascular
hypertension; IPF
+
= idiopathic
and extravascular components. Regional extravascular density (EVD) was calculated as TLD minus regional intravascular density (14). Lung regions were identified on transaxial slices of the transmission scan for each subject, stored in computer memory, and then superimposed on images of TLD, blood volume, EVD, and PTCER. These regions were defined on five to seven transverse slices (depending on patient positioning), representing as much as approximately 10.5 em of lung in a craniocaudal direction. Normal reference values for PTCER, TLD, and EVD for volunteer subjects, with no evidence of cardiopulmonary disease, are TLD, 39.0 ± 4.7 gm/loo mllung; EVD, 17.9 ± 3.6 g/loo mllung; PTCER, 21 ± 11 X 1O-4/min (8).
Seven recipients studied within 72 h of transplantation had a widened AaPo2 and infiltrates in the allograft (consistent with a reimplantation response) at the time of the PET study. The average PTCER for all slices from each of the 10 allografts in the Reimplantation Group are shown in figure IA as a function of the post-transplant day each PET scan was performed. In each case, PTCER was higher on the scan soon after surgery and decreased with time, although PTCER was actually normal in only one of these patients within the first 2 wk. Each of these subjects subsequently had an uncomplicated postoperative period with steadily improving oxygenation and chest radiographs. The PTCER values for each tomographic slice from these same 10 allografts are compared in figure IB with the duration of organ
preservation, measured as the time of organ harvest to the time of reperfusion after implantation. The preservation time correlated significantly with a single average value for each allograft (average initial values shown in figure IA). Data from the individual slices (individual symbols in figure IB), however, also show considerable variability, even within the same allograft. Three subjects in this group receivedbilateral transplants, in which each allograft was transplanted separately. Thus, the preservation times for each allograft were different, even in the same recipient (figure IC). In each case, the longer preservation time was associated with higher PTCER values and the correlation for all six allografts was r = 0.84. Images from one slice in a representative patient are shown in figure 2. PTCER is elevated on the first postoperative day and decreases by Day 4, even while lung density (comparable to infiltrates on the standard radiograph) increases. PTCER values for the 11 allografts (eight recipients) studied prior to transbronchial biopsy performed to diagnose rejection (Rejection Group) are shown in figure 3. Three subjects who wereclinicallystable (fiveallografts) underwent bronchoscopy for routine surveillance. In all five allografts, PTCER was normal, and no evidence for rejection was found in the transbronchial biopsies. Five patients (six allografts) underwent bronchoscopy for a clinical suspicion of rejection, which was confirmed histologically in each case. Each of these recipients had an abnormal average PTCER; however,regional values for PTCER in individual recipients varied widely (data not shown). Follow-up scans were performed in two of these five recipients (three allografts) after treatment for rejection with methylprednisolone (figure 3). In both cases, PTCER fell, and clinical signs of rejection subsided. Images from one serial PET study in this group are shown in figure 4. In this case, the regional heterogeneity in PTCER, even within a single sli~~ is evident. This subject underwent PET-guided bronchoscopy, in which tissue was obtained from areas of normal and abnormal PTCER defined by PET imaging on the previous day. Histologic examination later confirmed the physiologic findings. This patient also shows the response in PTCER after successful treatment for rejection.
* * ** Although pulmonary vascular permeability is classically elevated in the adult respiratory distress syndrome, recent studies from our group indicate that permeability is often abnormal in other lung injury states characterized by active inflammation (7,8, 10). However, since PTCER (the quantitative index of vascular permeability used in these studies) is abnormal during the adult respiratory distress syndrome and lobar pneumonia, but is normal during congestive heart failure (8), an elevated PTCER is apparently a sensitive, but nonspecific, index of lung injury or inflam-
956
BRIEF COMMUNICATION
225.-
225
175
A
-
200
150
175
e
! I 0
.::::.
ffi
~
CII
~
100
8
r = 0.77 p < 0.05 n - 10
I
•
•
•
50
25
0
0
-. 2
4
8
10
Post Transplant Day
175
*
125
..
150
zI
125
......
100
.:::.
100
75
0::: l.LJ 0 l-
75
c,
e
'E
a,
50
,.--."
c
•
.1
24 h after implantation. However, these three allografts also had the shortest ischemic times (table 1). Thus, the low values for PTCER in these three allografts could be due to either the short preservation time or the tendency of PTCER to decrease in general with time (figure IB). The diagnosis of allograft rejection, occurring later in the course after a transplant, has
957
BRIEF COMMUNICATION
Fig. 4. Representative PET images for a patient with clinical evidence for allograft rejection (orientation of images as in figure 2). Transbronchial lung biopsy was guided to an area of elevated PTCER (= 162) and was abnormal histologically in contrast to the region of normal PTCER (= 7). After treatment with increased immunosuppression, PTCER fell to normal values.
also been a difficult problem in management. In our first seven subjects studied for rejection, we used the standard clinical definition to diagnose rejection, which was then corroborated after histologic review oftransbronchial biopsy specimens. The PET scans were performed after the biopsies were obtained, but before the histologic results were known. As in the Reimplantation Group (figure IB), slice-to-slice and within-slice heterogeneity in PTCER was noted among the studies in the Rejection Group. Thus, in the eighth subject, the PET study was performed prior to bronchoscopy, and a PET-guided biopsy was performed. Tissue was obtained from an area of increased PTCER, which histologically demonstrated rejection, whereas biopsy specimens from other areas with a normal PleER showed no rejection (figure 4). This is consistent with the well-recognized spatial heterogeneity in allograft rejection found in lung biopsy studies that contributes to the current clinical practice of performing multiple transbronchial biopsies during each bronchoscopy for a diagnosis of rejection (27). When this eighth patient was studied by PET-guided biopsy after treatment for rejection, all regions had a normal PTCER, and no evidence of rejection was found on multiple biopsy specimens. In conclusion, these data suggest that PET might be an excellent tool to evaluate both the reimplantation response and allograft rejection after lung transplantation. The tomographic display format of PET images might allow one to guide biopsy procedures to optimal sites for histologic diagnoses, reducing the risks associated with the current practice of multiple biopsies. Because an elevation in PTCER is nonspecific as to its cause, PET studies would not be expected to eliminate the need to use bronchoscopy for definitive etiologic diagnosis. On the other hand, the possibility that PleER also changes in response to treatment for rejection might al-
low one to evaluate different immunosuppressive protocols. As to an evaluation of the reimplantation response, the good correlation between PleER and ischemic times prior to lung implantation in this study implies that PET studies could be used to evaluate new lung preservation protocols. Because current practice limits ischemic times to approximately 6 h, the reimplantation response has rarely limited the success of the lung transplant. However, as the relationship between reimplantation and ischemia-reperfusion injury is better defined, and as the need for longer ischemic times increases to expand the availability of donor lungs, the techniques described in this report may be extremely useful for evaluating new regimens designed to reduce injury after implantation.
References 1. Herman Sl, Weisbrod GL, Weisbrod L, Pat-
terson GA, Maurer lR. Chest radiographic findings after bilaterallung transplantation. AJR 1989; 153:1181-5. 2. Montefusco CM, Veith Fl. Lung transplantation. Surg Clin North Am 1986; 66:305-15. 3. Herman Sl. Radiologic assessment after lung transplantation. Clin Chest Med 1990; 11:333-46. 4. Herman Sl, Rappaport DC, Weisbrod GL, Olscamp C, Patterson GA, Cooper 10. Single-lung transplantations:imagingfeatures. Radiology1989; 170:89-93. 5. Millet B, Higenbottam TW, Flower CDR, Stewart S, Wallwork J. The radiographic appearances of infection and acute rejection of the lung after heart-lung transplantation. Am Rev Respir Dis 1989; 140:62-7. 6. Schuster DP. Positron emission tomography: theory and its application to the study of lung disease. Am Rev Respir Dis 1989; 139:818-40. 7. Calandrino FS, Anderson Dl, Mintun MA, SchusterDP. Pulmonary vascularpermeabilityduring the adult respiratory distress syndrome: a positron emission tomographic study. Am Rev Respir Dis 1988; 138:421-8. 8. Kaplan JD, Calandrino FS, Schuster DP. A
positron emissiontomographic comparison of pulmonary vascular permeability during the adult respiratory distress syndrome and pneumonia. Am Rev Respir Dis 1990; 143:150-4. 9. Kaplan lD, Calandrino FS, Schuster DP. Pulmonary vascular permeability as measured by positron emissiontomography is unaffected bycigarette smoking (abstract). Am RevRespirDis 1989; 141:A424. 10. KaplanJD, TrulockEP,AndersonDl, Schuster DP. Pulmonary vascular permeability in interstitiallung disease:a positron emission tomographic study. Am Rev Respir Dis 1992; (In Press). 11. Ter-Pogossian MM, Ficke DC, YamomotoM, Hood IT. Design characteristics and preliminary testing of SUPERPETT I, a positron emission tomography utilizing photon time-of-flight information (TOF-PET). IEEE Trans Med Imag 1982; 1:37-42. 12. Mintun MA, Dennis DR, WelchMF, Mathias Cl, SchusterDP. Measurementsof pulmonary vascular permeabilitywith PET and gallium-68transferrin. J Nucl Med 1987; 28:1704-16. 13. Dittmer DS. Respiration and circulation. Bethesda, MD: FASEB, 1971; 15-20. 14. Schuster DP, Marklin GF, Mintun MA, TerPogossianMM. PET measurementof regionallung density. 1 Comput Assist Tomogr 1986; 10:723-9. 15. ReichartBA, HumanPA, RoseAG, Novitzky D, Cooper DKC. Is pulmonary ischemia a factor in reperfusion response?An experimental study in the chacma baboon. J Heart Transplant 1987; 6:238-43. 16. KoernerSK, Hagstrom lWC, Veith FJ. Transbronchial biopsy for the diagnosis of lung transplant rejection.Am RevRespirDis 1976; 114:575-9. 17. Cowan GSM, Staub NC, Edmunds LH. Changes in the fluidcompartments and dry weights of reimplanted dog lungs. 1 Appl Physiol 1976; 40:962-70. 18. Halasz NA, Catanzaro A, Trummer 1M, et al. Transplantation of the lung. Correlation of physiologic, immunologic, and histologic findings. J Thorac Cardiovasc Surg 1973; 66:581-7. 19. Horgan MJ, Lum H, Malik AB. Pulmonary edema after pulmonary artery occlusionand reperfusion. Am Rev Respir Dis 1989; 140:1421-8. 20. Groningen JM, Ehrie MG, Crapo JD, Nieuwenhuis P, Wildevuur CRH. Reimplantation response in isografted rat lungs. Analysisof causal factors. 1 Thorac CardiovascSurg 1984; 87:702-11. 21. Siegelman SS, Hagstrom JW, Veith FJ. Roentgenologic-pathologic evaluation of rejection in allografted lungs. AJR 1971; 112:546-58. 22.. Eraslan S, Turner M, Hardy Jf). Lymphatic regenerationfollowing lung reimplantation in dogs. Surgery 1964; 56:970-3. 23. Kehavjee SH, Herman SJ, Yamazaki F,Slutsky AS, Cooper JD, Patterson GA. Radiologic correlation with early physiologicfunction of the transplanted canine lung. Invest Radio11990; 25:511-6. 24. SnashallPD, Keyes SJ, Morgan BM, et al. The radiographic detectionof acute pulmonary edema. A comparisonof radiologicappearances, densitometry and lung water in dogs. Br 1 Radiol 1981; 54:277-88. 25. Hruban RJ, Ren H, Kuhlman E, et al. Inflation-fixed lungs: pathologic-radiologic (CT) correlation oflung transplantation. J Comput Assist Tomogr 1990; 14:329-35. 26. Starnes VA, Theodore 1, Oyer PE, et al. Pulmonary infiltratesafter heart-lung transplantation: evaluation by serial transbronchial biopsies. J Thorac Cardiovasc Surg 1989; 98:945-50. 27. StarnesVA, Theodore1, OyerPE, et al. Evaluation of heart-lung transplant recipients with prospective, serialtransbronchial biopsies and pulmonary function studies.1 Thorac CardiovascSurg 1989; 98:683-90.
JAMES D. KAPLAN, ELBERT P. TRULOCK, JOEL D. COOPER, and DANIEL P. SCHUSTER4
During the past 10yr, human lung transplantation has rapidly become a valuable therapeutic option for end-stage lung disease. Among many important issues, the so-called "reimplantation response" and allograft rejection stand out as two processes of particular importance. Although both are characterized by histologic evidence of lung injury, each process has a distinctive clinical course (1). The reimplantation response has been defined as the "morphologic, roentgenographic, and functional changes that occur in a lung transplant in the early postoperative period as a result of surgicaltrauma, ischemia, denervation, lymphatic interruption, and other injurious processes (exclusiveof rejection)" (2). This response is usually diagnosed by exclusion, i.e., when basal or perihilar infiltrates occur in the early postoperative period and cannot be attributed to rejection, infection, mucus plugging, or cardiogenic pulmonary edema. The severity of the response usually reaches a maximum by Day 4, and clears rapidly thereafter (3). The diagnosis is often confirmed retrospectively when the observed abnormalities resolve without specific therapy. In contrast, allograft rejection rarely occurs during these first few days after transplantation. After the first week, rejection is suspected when fever, dyspnea, malaise, leukocytosis, hypoxia, or new infiltrates occur, and the diagnosis is confirmed clinically by a prompt response to intravenously administered corticosteroids and/or histologically by a perivascular lymphocytic infiltrate in transbronchiallung biopsies (4). Interestingly, surveillance transbronchial biopsy studies after lung transplantation indicate that unsuspected allograft rejection can frequently be present without clinical signs and symptoms (5). Despite these distinctive features, these two complications of transplantation have been remarkably difficult to evaluate clinically using currently available noninvasive methods (1). We have developed a means to noninvasively evaluate pulmonary vascular permeability using the quantitative nuclear medicine imaging technique of positron emission tomography (PET) (6). This technique has been used to evaluate lung injury as a result of the adult respiratory distress syndrome (7), lobar pneumonia (8), chronic cigarette smoke exposure (9), and interstitial lung disease (10). Because 954
SUMMARY Weevaluated pulmonary vascular permeability In 15patients after lung transplantation (21 allografts) by measuring the pUlmonary transcapiilary escape rate (PTCER) for Ga·68·labeled transferrin, using positron emission tomography. Seven recipients (four unilateral, three bilateral lung transplants) were studied within 3 days of transplantation, and each developed hypoxemia and allograft Infiltrates consistent with the "relmplantatlon response." PTCER was higher In subjects studied within 1 day than In those studied at a later time, and fell In seven allografts studied serially. The initial PTCER also correlated (r = 0.77) with length of Ischemic (preservation) time, even In the three subjects with bilateral allografts. Eight other recipients (five unilateral, three bilateral transplants) were evaluated for possible organ rejection at least 1 wk after transplantation. PTCER was normal In patients without clinical or histologic evidence of rejection, and It was elevated In recipients with rejection. PTCER fell each time after treatment for rejection with Increased Immunosuppression In the three patients studied serially. These data suggest that positron emission tomography measurements of PTCER might be a useful way to evaluate both the relmplantation response and organ rejection after lung transplantation. AM REV RESPIR DIS 1992; 145:954-957
both the reimplantation response and allograft rejection are characterized by histologic and radiographic evidence oflung injury, weused PET to determine whether changes in pulmonary vascular permeability could be detected during either condition and whether these changes correlated with successful treatment. Twenty-threePET scans wereperformed on 15adult single- and double-lung transplant recipients. Age, sex, indication for transplantation, and other clinical parameters are indicated in tables 1 and 2. The protocol for this study was approved by the Washington University Human Studies Committee. To evaluate the reimplantation response (defined operationally for this study as hypoxemia requiring supplemental oxygen and new radiographic infiltrates in the allograft), PET scans were performed within 72 h of surgery on seventransplant recipients who were stable for transportation to the scanner. Three of the seven recipients had bilateral transplants, so that a total of 10 lung allografts were studied (Reimplantation Group). Serial scans were repeated in five recipients (seven of 10 allografts) 4 to 8 days later. Eleven allografts in eight other recipients were studied at least 1wk after transplantation to evaluate pulmonary vascular permeability during rejection (Rejection Group). PET scans were performed within the 24 h before subjects underwent bronchoscopy and transbronchial biopsy, either for routine rejection surveillance or for a clinical suspicion of rejection. Histologic evidence of a perivascular mononuclear cell infiltrate with negative microbiologic cultures was considered diagnostic for rejection. Serial studies were performed in three patients after treatment for rejection with intravenously administered methylprednisolone.
PET scan data were acquired on the SUPERPETT I scanner (11),and displayed as seven transaxial tomographic slices with a center-to-center separation of 15mm and an intrinsic in-plane resolution of 18 mm. The methods and theory of PET measurements have been reviewed recently (3). Each PET study included a background scan, a transmission scan, and a series of emission scans after the injection of radioisotope. Sternal landmarks were used to guide patient positioning in the scanner, and these same landmarks were used in the case of serial studies in the same patient. The background and transmission scans wereperformed to correct subsequent emission scans for errors introduced by tissue attenuation during the tissue activity measurement. Because photon attenuation is proportional to tissue density, the transmission
(Received in original form August 2, 1991 and in revised form September 27, 1991) 1 From the Respiratory and Critical Care Division, Department of Internal Medicine, Washington University School of Medicine, St. Louis, Missouri. 2 Supported by Grants R01 HL-41943 and POI HL-13851 from the National Institutes of Health and by Grant DE-FG02-87ER60512 from the Department of Energy. 3 Correspondence and requests for reprints should be addressed to Daniel P. Schuster, M.D., Washington University School of Medicine, Respiratory and Critical Care Division, 660 S. Euclid, Box 8052, St. Louis, MO 63110. 4 Established Investigator of the American Heart Association and a Career Investigator of the American Lung Association.
955
BRIEF COMMUNICATION
TABLE 1 REIMPLANTATION RESPONSE STUDIES Patient No. 1 2 3 4 5 5 6 6 7 7
(yr)
Sex
Side
Indication for Transplant
Ischemic Time
Hours from Reperfusion to Scan
Hours to Follow-up Scan
61 22 53 35 21 21 42 42 53 53
M F F F F F M M F F
L L L R L R L R L R
COPD Bleo Fibrosis COPD EG CF CF AAT AAT Bronchiectasis Bronchiectasis
3' 32" 3' 50" 5' 20" 3' 30" 7'17" 5' 20" 6' 0" 4' 16" 4' 35" 7' 9"
53 70 10 41 24 24 34 34 27 27
238 251 108
Age
118 118 124 124
Definition of abbreviations: L = left; R = right; COPO = chronic obstructive pulmonary disease; Bleo sinophilic granuloma; CF = cystic fibrosis; AAT = alpha-1-antitrypsin deficiency.
= bleomycin;
EG
= Eo-
TABLE 2 REJECTION STUDIES Patient No. 1 2 3 4 5 6 7 8
Age (yr)
Sex
Side
Indication
47 57 46 53 35 37 28 35
F F F F F F F F
R L R L B R B B
COPD COPD COPD COPD COPD PPH IPF AAT
Definition of abbreviations: POD = postoperative day; B pulmonary fibrosis. For other definitions, see table 1.
Rejection Suspected
POD for Initial Scan
Follow-up Performed
+ +
37 7 17 58 30 190 358 104
+
+ + +
= both; PPH = preliminary pulmonary
scan also provides a density image, resembling a standard X-ray-computed tomographic image, which can be used to define anatomic regions of interest (12). Radioisotope was prepared according to previously described methods (12). As much as 6 mCi of 68Ga-citrate were injected intravenously. After injection, 68Ga-citrate rapidly dissociates and the 68Ga then avidly binds to circulating transferrin, serving both as a plasma marker initially and subsequently as a tracer for the measurement of protein flux into the extravascular space. Serial emission scans were collected for 60 min beginning 2 min after injection. Time-activity curves were measured for the PET activity in the lung regions of interest and within the cardiac blood pool on each emission scan and decay-corrected to the time of tracer injection. The blood pool activity was analyzed by two sequential linear regressions to determine the best fit as either a monoexponential or biexponential expression for the blood time-activity curve (12). The tissue and blood time-activity curves were then analyzed according to a two-compartment mathematical model (8, 12). Using this model, a vascular-to-extravascular rate constant (K1) was calculated. Because protein flux is affected by surface area available for exchange, K 1 was normalized to regional blood volume, which was assumed to be proportional to surface area. Regional blood volume was calculated from the initial 2-min (68Ga]transferrin scan, when virtually all tracer activity was still intravascular. Regional pulmonary transcapillary escape rate (PTCER) was then expressed as K1/blood volume. Lung density was calculated by scaling the attenuation scan to the known density of blood, so that the cardiac blood pool had a density of 1.05 g/ml (13). This resulted in images of regional total lung density (TLD), which included both vascular
hypertension; IPF
+
= idiopathic
and extravascular components. Regional extravascular density (EVD) was calculated as TLD minus regional intravascular density (14). Lung regions were identified on transaxial slices of the transmission scan for each subject, stored in computer memory, and then superimposed on images of TLD, blood volume, EVD, and PTCER. These regions were defined on five to seven transverse slices (depending on patient positioning), representing as much as approximately 10.5 em of lung in a craniocaudal direction. Normal reference values for PTCER, TLD, and EVD for volunteer subjects, with no evidence of cardiopulmonary disease, are TLD, 39.0 ± 4.7 gm/loo mllung; EVD, 17.9 ± 3.6 g/loo mllung; PTCER, 21 ± 11 X 1O-4/min (8).
Seven recipients studied within 72 h of transplantation had a widened AaPo2 and infiltrates in the allograft (consistent with a reimplantation response) at the time of the PET study. The average PTCER for all slices from each of the 10 allografts in the Reimplantation Group are shown in figure IA as a function of the post-transplant day each PET scan was performed. In each case, PTCER was higher on the scan soon after surgery and decreased with time, although PTCER was actually normal in only one of these patients within the first 2 wk. Each of these subjects subsequently had an uncomplicated postoperative period with steadily improving oxygenation and chest radiographs. The PTCER values for each tomographic slice from these same 10 allografts are compared in figure IB with the duration of organ
preservation, measured as the time of organ harvest to the time of reperfusion after implantation. The preservation time correlated significantly with a single average value for each allograft (average initial values shown in figure IA). Data from the individual slices (individual symbols in figure IB), however, also show considerable variability, even within the same allograft. Three subjects in this group receivedbilateral transplants, in which each allograft was transplanted separately. Thus, the preservation times for each allograft were different, even in the same recipient (figure IC). In each case, the longer preservation time was associated with higher PTCER values and the correlation for all six allografts was r = 0.84. Images from one slice in a representative patient are shown in figure 2. PTCER is elevated on the first postoperative day and decreases by Day 4, even while lung density (comparable to infiltrates on the standard radiograph) increases. PTCER values for the 11 allografts (eight recipients) studied prior to transbronchial biopsy performed to diagnose rejection (Rejection Group) are shown in figure 3. Three subjects who wereclinicallystable (fiveallografts) underwent bronchoscopy for routine surveillance. In all five allografts, PTCER was normal, and no evidence for rejection was found in the transbronchial biopsies. Five patients (six allografts) underwent bronchoscopy for a clinical suspicion of rejection, which was confirmed histologically in each case. Each of these recipients had an abnormal average PTCER; however,regional values for PTCER in individual recipients varied widely (data not shown). Follow-up scans were performed in two of these five recipients (three allografts) after treatment for rejection with methylprednisolone (figure 3). In both cases, PTCER fell, and clinical signs of rejection subsided. Images from one serial PET study in this group are shown in figure 4. In this case, the regional heterogeneity in PTCER, even within a single sli~~ is evident. This subject underwent PET-guided bronchoscopy, in which tissue was obtained from areas of normal and abnormal PTCER defined by PET imaging on the previous day. Histologic examination later confirmed the physiologic findings. This patient also shows the response in PTCER after successful treatment for rejection.
* * ** Although pulmonary vascular permeability is classically elevated in the adult respiratory distress syndrome, recent studies from our group indicate that permeability is often abnormal in other lung injury states characterized by active inflammation (7,8, 10). However, since PTCER (the quantitative index of vascular permeability used in these studies) is abnormal during the adult respiratory distress syndrome and lobar pneumonia, but is normal during congestive heart failure (8), an elevated PTCER is apparently a sensitive, but nonspecific, index of lung injury or inflam-
956
BRIEF COMMUNICATION
225.-
225
175
A
-
200
150
175
e
! I 0
.::::.
ffi
~
CII
~
100
8
r = 0.77 p < 0.05 n - 10
I
•
•
•
50
25
0
0
-. 2
4
8
10
Post Transplant Day
175
*
125
..
150
zI
125
......
100
.:::.
100
75
0::: l.LJ 0 l-
75
c,
e
'E
a,
50
,.--."
c
•
.1
24 h after implantation. However, these three allografts also had the shortest ischemic times (table 1). Thus, the low values for PTCER in these three allografts could be due to either the short preservation time or the tendency of PTCER to decrease in general with time (figure IB). The diagnosis of allograft rejection, occurring later in the course after a transplant, has
957
BRIEF COMMUNICATION
Fig. 4. Representative PET images for a patient with clinical evidence for allograft rejection (orientation of images as in figure 2). Transbronchial lung biopsy was guided to an area of elevated PTCER (= 162) and was abnormal histologically in contrast to the region of normal PTCER (= 7). After treatment with increased immunosuppression, PTCER fell to normal values.
also been a difficult problem in management. In our first seven subjects studied for rejection, we used the standard clinical definition to diagnose rejection, which was then corroborated after histologic review oftransbronchial biopsy specimens. The PET scans were performed after the biopsies were obtained, but before the histologic results were known. As in the Reimplantation Group (figure IB), slice-to-slice and within-slice heterogeneity in PTCER was noted among the studies in the Rejection Group. Thus, in the eighth subject, the PET study was performed prior to bronchoscopy, and a PET-guided biopsy was performed. Tissue was obtained from an area of increased PTCER, which histologically demonstrated rejection, whereas biopsy specimens from other areas with a normal PleER showed no rejection (figure 4). This is consistent with the well-recognized spatial heterogeneity in allograft rejection found in lung biopsy studies that contributes to the current clinical practice of performing multiple transbronchial biopsies during each bronchoscopy for a diagnosis of rejection (27). When this eighth patient was studied by PET-guided biopsy after treatment for rejection, all regions had a normal PTCER, and no evidence of rejection was found on multiple biopsy specimens. In conclusion, these data suggest that PET might be an excellent tool to evaluate both the reimplantation response and allograft rejection after lung transplantation. The tomographic display format of PET images might allow one to guide biopsy procedures to optimal sites for histologic diagnoses, reducing the risks associated with the current practice of multiple biopsies. Because an elevation in PTCER is nonspecific as to its cause, PET studies would not be expected to eliminate the need to use bronchoscopy for definitive etiologic diagnosis. On the other hand, the possibility that PleER also changes in response to treatment for rejection might al-
low one to evaluate different immunosuppressive protocols. As to an evaluation of the reimplantation response, the good correlation between PleER and ischemic times prior to lung implantation in this study implies that PET studies could be used to evaluate new lung preservation protocols. Because current practice limits ischemic times to approximately 6 h, the reimplantation response has rarely limited the success of the lung transplant. However, as the relationship between reimplantation and ischemia-reperfusion injury is better defined, and as the need for longer ischemic times increases to expand the availability of donor lungs, the techniques described in this report may be extremely useful for evaluating new regimens designed to reduce injury after implantation.
References 1. Herman Sl, Weisbrod GL, Weisbrod L, Pat-
terson GA, Maurer lR. Chest radiographic findings after bilaterallung transplantation. AJR 1989; 153:1181-5. 2. Montefusco CM, Veith Fl. Lung transplantation. Surg Clin North Am 1986; 66:305-15. 3. Herman Sl. Radiologic assessment after lung transplantation. Clin Chest Med 1990; 11:333-46. 4. Herman Sl, Rappaport DC, Weisbrod GL, Olscamp C, Patterson GA, Cooper 10. Single-lung transplantations:imagingfeatures. Radiology1989; 170:89-93. 5. Millet B, Higenbottam TW, Flower CDR, Stewart S, Wallwork J. The radiographic appearances of infection and acute rejection of the lung after heart-lung transplantation. Am Rev Respir Dis 1989; 140:62-7. 6. Schuster DP. Positron emission tomography: theory and its application to the study of lung disease. Am Rev Respir Dis 1989; 139:818-40. 7. Calandrino FS, Anderson Dl, Mintun MA, SchusterDP. Pulmonary vascularpermeabilityduring the adult respiratory distress syndrome: a positron emission tomographic study. Am Rev Respir Dis 1988; 138:421-8. 8. Kaplan JD, Calandrino FS, Schuster DP. A
positron emissiontomographic comparison of pulmonary vascular permeability during the adult respiratory distress syndrome and pneumonia. Am Rev Respir Dis 1990; 143:150-4. 9. Kaplan lD, Calandrino FS, Schuster DP. Pulmonary vascular permeability as measured by positron emissiontomography is unaffected bycigarette smoking (abstract). Am RevRespirDis 1989; 141:A424. 10. KaplanJD, TrulockEP,AndersonDl, Schuster DP. Pulmonary vascular permeability in interstitiallung disease:a positron emission tomographic study. Am Rev Respir Dis 1992; (In Press). 11. Ter-Pogossian MM, Ficke DC, YamomotoM, Hood IT. Design characteristics and preliminary testing of SUPERPETT I, a positron emission tomography utilizing photon time-of-flight information (TOF-PET). IEEE Trans Med Imag 1982; 1:37-42. 12. Mintun MA, Dennis DR, WelchMF, Mathias Cl, SchusterDP. Measurementsof pulmonary vascular permeabilitywith PET and gallium-68transferrin. J Nucl Med 1987; 28:1704-16. 13. Dittmer DS. Respiration and circulation. Bethesda, MD: FASEB, 1971; 15-20. 14. Schuster DP, Marklin GF, Mintun MA, TerPogossianMM. PET measurementof regionallung density. 1 Comput Assist Tomogr 1986; 10:723-9. 15. ReichartBA, HumanPA, RoseAG, Novitzky D, Cooper DKC. Is pulmonary ischemia a factor in reperfusion response?An experimental study in the chacma baboon. J Heart Transplant 1987; 6:238-43. 16. KoernerSK, Hagstrom lWC, Veith FJ. Transbronchial biopsy for the diagnosis of lung transplant rejection.Am RevRespirDis 1976; 114:575-9. 17. Cowan GSM, Staub NC, Edmunds LH. Changes in the fluidcompartments and dry weights of reimplanted dog lungs. 1 Appl Physiol 1976; 40:962-70. 18. Halasz NA, Catanzaro A, Trummer 1M, et al. Transplantation of the lung. Correlation of physiologic, immunologic, and histologic findings. J Thorac Cardiovasc Surg 1973; 66:581-7. 19. Horgan MJ, Lum H, Malik AB. Pulmonary edema after pulmonary artery occlusionand reperfusion. Am Rev Respir Dis 1989; 140:1421-8. 20. Groningen JM, Ehrie MG, Crapo JD, Nieuwenhuis P, Wildevuur CRH. Reimplantation response in isografted rat lungs. Analysisof causal factors. 1 Thorac CardiovascSurg 1984; 87:702-11. 21. Siegelman SS, Hagstrom JW, Veith FJ. Roentgenologic-pathologic evaluation of rejection in allografted lungs. AJR 1971; 112:546-58. 22.. Eraslan S, Turner M, Hardy Jf). Lymphatic regenerationfollowing lung reimplantation in dogs. Surgery 1964; 56:970-3. 23. Kehavjee SH, Herman SJ, Yamazaki F,Slutsky AS, Cooper JD, Patterson GA. Radiologic correlation with early physiologicfunction of the transplanted canine lung. Invest Radio11990; 25:511-6. 24. SnashallPD, Keyes SJ, Morgan BM, et al. The radiographic detectionof acute pulmonary edema. A comparisonof radiologicappearances, densitometry and lung water in dogs. Br 1 Radiol 1981; 54:277-88. 25. Hruban RJ, Ren H, Kuhlman E, et al. Inflation-fixed lungs: pathologic-radiologic (CT) correlation oflung transplantation. J Comput Assist Tomogr 1990; 14:329-35. 26. Starnes VA, Theodore 1, Oyer PE, et al. Pulmonary infiltratesafter heart-lung transplantation: evaluation by serial transbronchial biopsies. J Thorac Cardiovasc Surg 1989; 98:945-50. 27. StarnesVA, Theodore1, OyerPE, et al. Evaluation of heart-lung transplant recipients with prospective, serialtransbronchial biopsies and pulmonary function studies.1 Thorac CardiovascSurg 1989; 98:683-90.
