tions of TGF- 2 at doses of 5.00 g (n 12), 1.67 g (n 10), 0.50 g (n 10), or 0.167 ...... Light, R. W., N. S. Wang, C. S. H. Sassoon, S. E. Gruer, and F. S. Vargas. 1994.
A Single Intrapleural Injection of Transforming Growth Factor-2 Produces an Excellent Pleurodesis in Rabbits RICHARD W. LIGHT, DONG-SHENG CHENG, Y. C. (GARY) LEE, JEFFREY ROGERS, JEFFREY DAVIDSON, and KIRK B. LANE Department of Medicine, Saint Thomas Hospital, and Center for Lung Research and Pathology, Vanderbilt University, Nashville, Tennessee
The purpose of the present study was to determine whether the intrapleural injection of transforming growth factor 2 (TGF-2) would produce a pleurodesis in rabbits. Single intrapleural injections of TGF-2 at doses of 5.00 g (n ⫽ 12), 1.67 g (n ⫽ 10), 0.50 g (n ⫽ 10), or 0.167 g (n ⫽ 4), or of the parenteral buffer alone (n ⫽ 5) were given in a volume of 2 ml to New Zealand white rabbits. Chest tubes were left in place for at least 72 h. Pleural fluid was aspirated at 24-h intervals and was measured and subjected to chemical analysis. The animals were killed 14 d after the injection. The intrapleural injection of TGF-2 resulted in a dose-dependent pleurodesis (on a scale of 0 to 4, where 0 ⫽ no pleurodesis and 4 ⫽ complete pleurodesis) with mean scores of 3.6, 2.6, 1.5, 0.7, and 0.3 for the groups that received 5.0, 1.67, 0.50, and 0.167 g of TGF-2 and buffer alone, respectively. Intrapleural injection of the larger doses of TGF-2 resulted in the formation of a large amount of pleural fluid. The fluid had a significantly lower white blood cell (WBC) count and lactate dehydrogenase (LDH) level than did the fluid that results from the intrapleural injection of 10 mg/kg doxycycline or 400 mg/kg talc slurry. On the basis of this study we conclude that a single intrapleural injection of TGF-2 induces pleurodesis in a dose-dependent manner. A dose of 5.0 g produced satisfactory pleurodesis in almost all of the rabbits so treated. Larger doses of TGF-2 induced larger pleural effusions with relatively low pleural fluid WBC counts and LDH levels. The ability of TGF- to produce a pleurodesis in patients with recurrent pleural effusions or pneumothorax should be investigated. A single intrapleural injection of TGF-2 may produce a pleurodesis both safely and painlessly.
The treatment of recurrent pleural effusion or recurrent pneumothorax frequently involves the creation of a pleurodesis. With a pleurodesis, fusion of the visceral and parietal pleurae occurs, leaving no space for the collection of pleural fluid or air. In clinical practice, a pleurodesis is most commonly created by injecting a sclerosing agent through a chest tube. The agents most commonly used are antineoplastic drugs (bleomycin, mitoxantrone, or nitrogen mustard), tetracycline derivatives (doxycycline or minocycline), or talc slurry. None is ideal. Bleomycin is relatively expensive (ⵑ $1,000/patient) and is probably less efficacious than are the tetracycline derivatives or talc (1). The injection of the tetracycline derivatives is at times extremely painful (2). The injection of talc can lead to the development of acute respiratory distress syndrome and death (3, 4). Accordingly, the search continues for an ideal sclerosing agent. The exact mechanisms responsible for pleurodesis are unknown. It is thought that the intrapleural injection of an irritant produces inflammation (5), which in turn initiates a cas-
(Received in original form September 27, 1999 and in revised form December 22, 1999 ) Supported by the Saint Thomas Foundation, Nashville, TN, and Genzyme Corporation, Framingham, MA. Correspondence and requests for reprints should be addressed to Richard W. Light, M.D., Director of Pulmonary Disease Program, Saint Thomas Hospital, P.O. Box 380 - 4220 Harding Rd., Nashville, TN 37205. E-mail: rlight98@yahoo. com Am J Respir Crit Care Med Vol 162. pp 98–104, 2000 Internet address: www.atsjournals.org
cade of events that culminates at times in the creation of a pleurodesis. Cytokines are unquestionably involved in the processes that result in pleurodesis. Theoretically, it might be possible to create a pleurodesis by the intrapleural injection of a cytokine, which would avoid the need for injury to the pleura. If a solitary cytokine could produce a pleurodesis, the following characteristics of transforming growth factor  (TGF-) make it a good candidate cytokine for this purpose: (1) TGF- is a potent fibrogenic cytokine that regulates extracellular matrix (ECM) production; in situations in which there is too much TGF-, fibrosis results (6). The transient overexpression of TGF- in the rat lung leads to marked pleural and interstitial fibrosis (7). (2) Once present, TGF- can induce its own transcription (8), which suggests that a single injection may be sufficient to induce pleurodesis. (3) Mesothelial cells express and secrete TGF-; therefore, a single intrapleural injection of TGF- might result in prolonged secretion of TGF-, which could result in pleurodesis. (4) The incubation of human pleural mesothelial cells with TGF- results in secretion of increased levels of plasminogen activator inhibitor (PAI)-1 (9), which could facilitate pleurodesis, since inhibition of the fibrinolytic system is thought to be necessary for the production of a pleurodesis (10). The purpose of the present study was to determine whether the intrapleural injection of TGF- could result in a pleurodesis. We hypothesized that the intrapleural injection of TGF- would result in a pleurodesis and would produce less inflammation than is produced by the usual agents used for pleurodesis.
METHODS The methods used were similar to those previously described in our rabbit model of pleurodesis (11–13). The study protocol was approved by the Animal Care Committee of the Vanderbilt University Medical Center. New Zealand white male rabbits weighing 1.5 to 2.5 kg were lightly anesthetized with ketamine hydrochloride (35 mg/kg) plus xylazine hydrochloride (5 mg/kg) intramuscularly. The thorax was prepared for aseptic surgery by shaving the right chest wall and then sterilizing it with povidone–iodine and alcohol. The rabbit was placed in the lateral decubitus position. A 0.5-cm skin incision was made at the right lateral chest wall, approximately 4 cm anterior to the spine and 2 cm above the costal margin. The parietal pleura was exposed by bluntly dissecting the muscles underlying the skin incision. The parietal pleura was then punctured. This allowed the right lung to collapse, thereby preventing damage to the lung when a catheter was inserted. A soft silastic catheter (I.D. ⫽ 1/32 in., O.D. ⫽ 1/16 in.; Dow Corning, Midland, MI) was inserted into the pleural space. In sequence, the muscles and tissues were closed with sutures. The proximal tip of the catheter was tunneled under the skin, drawn out through the skin posteriorly and superiorly between the two scapulae, and fixed to the skin with a silk suture. The exterior end of the catheter was plugged with a stub adaptor having a Luer-lock fitting (Cole Parmer, Vernon Hill, IL). A self-sealing injection-site fitting, with a Luer lock, was then attached to the stub adaptor (Baxter Health Care Corporation, Deerfield, IL). Air or liquid could be aspirated from the pleural space by inserting an 18gauge needle through the self-sealing plug. After the surgery, the rabbits were closely monitored for clinical evidence of pain (vocalization,
Light, Cheng, Lee, et al.: TGF- for Pleurodesis
99
tachypnea, and restlessness). The left hemithorax received no chest tube and no injection, and served as a control. The chest tubes were left in place for 24 h before any intrapleural injections were given. The primary goal of this research was to evaluate whether pleurodesis could be produced by the intrapleural injection of TGF-. The TGF- used for these experiments was the recombinant human TGF2 isoform (Genzyme Corporation, Framingham, MA), which is produced in Chinese hamster ovarian cells and is formulated in a vehicle consisting of 20 mM sodium phosphate, 130 mM sodium chloride, 15% (wt/wt) propylene glycol, and 20% (wt/wt) polyethylene glycol 400, pH 7.2. The vehicle was prepared with United States Pharmacopoeia/National Formulary-grade reagents in water for injection, and was sterile-filtered through a 0.2-m filter. The stability of TGF-2 in the vehicle had previously been demonstrated (Genzyme Corporation, unpublished data). The TGF-2 concentration was determined with a sandwich-type enzyme-linked immunosorbent assay (ELISA), utilizing two monoclonal antibodies that cross-react with TGF-2 and TGF-3. The activity of the TGF-2 was determined with a mink lung cell (Mv1Lu) antiproliferation assay, modified from the method described by Ogawa and Seyedin (14). Two separate studies are included in this report. The initial study was a pilot study to determine whether the intrapleural injection of TGF-2 could produce a pleurodesis. The second and primary study was a dose–response study of the production of pleurodesis by single doses of TGF-2. Ten rabbits were included in the pilot study (Table 1). The first two rabbits served as controls, and respectively received one and four injections of the buffer alone. Two rabbits received one and four injections, respectively, of TGF-2 in a dose of 5 g, two rabbits received one and four injections of TGF-2 at 10 g; and two rabbits received one and four injections of TGF-2 at 10 g. Two further rabbits received two and three injections, respectively, of 20 g of TGF-2. When rabbits received multiple injections, the injections were separated by 24 h. All injections had a total volume of 2 ml, and were followed by injection of 1.0 ml of 0.9% sodium chloride to clear the dead space of the chest tube. In the dose–response study, rabbits received single intrapleural injections of TGF-2 in doses of 0.167 g (n ⫽ 4), 0.50 g (n ⫽ 10), 1.67 g (n ⫽ 10), or 5 g (n ⫽ 12), or received buffer alone (n ⫽ 5). The total volume of each injection was 2 ml. The pleural space was aspirated via the chest tube every 24 h after the initial pleural injection. The chest tube was left in place for 96 h following the initial pleural injection, or until the volume of pleural fluid aspirated over the preceding 24 h was less than 3 ml, whichever occurred later. The total volume of pleural fluid aspirated was recorded. In the pilot study, animals scheduled to have an additional intrapleural instillation had aspiration done immediately before the instillation. The levels of glucose, protein, and lactate dehydrogenase (LDH) in pleural fluid were determined with a Vitros Model 950 automated analyzer (Johnson & Johnson, Rochester, NY). The white blood cell (WBC) count in pleural fluid was made with a hemocytometer. Leukocyte differential counts were determined by manually counting 100 cells in a Wright-stained smear.
Rabbits were killed at 14 d by intravenous injection of pentobarbital via the marginal ear vein. The thorax was removed en bloc. The trachea was intubated with a small silastic catheter and the lungs were inflated with 10% formalin in phosphate buffered saline. After the inflation, the trachea was ligated with plastic ties and the entire thorax was submerged in a 10% formalin solution for at least 48 h. Necropsy was performed by two of the investigators (D.S.C. and R.W.L.), who were blinded with regard to the treatment of each animal. Each pleural cavity was exposed by making bilateral incisions through the diaphragms and through all the ribs at the midclavicular line. In this manner, the sternum and the medial portions of the anterior ribs were removed, so that the lung and pleural cavities could be evaluated. The presence or absence of hemothorax (clotted blood in the pleural cavity), and the position of the mediastinum in each animal, were recorded. The degree of pleurodesis observed grossly was graded according to the following scheme: 0 ⫽ normal pleural space; 1 ⫽ one to three small adhesions in the pleural space; 2 ⫽ ⬎ 3 scattered adhesions, but lung separates easily from chest wall; 3 ⫽ generalized, scattered adhesions, with areas where lung can be separated from chest wall only with difficulty; 4 ⫽ complete obliteration of pleural space by adhesion. The degree of pleurodesis and the characteristics of the pleural fluid after administration of 5 g TGF- were compared with results obtained after the intrapleural administration of 400 mg/kg talc or 10 mg/kg doxycycline, which have been reported previously (15). The protocol that had been followed in rabbits given doxycycline and talc was essentially identical to that which we used for our single-dose TGF- injections.
Statistical Analysis All data are expressed as mean ⫾ SEM unless otherwise stated. The pleurodesis scores and the pleural fluid volume, LDH concentration, WBC counts, and differential cell counts in the different groups were compared through two-way repeated-measures analysis of variance (ANOVA). The means in the various groups were compared through use of the Student–Newman–Keuls method. If the data failed tests of normality or equal variance, the medians were compared through Kruskal–Wallis one-way ANOVA on ranks. The medians were compared through Dunn’s method (Sigma Stat; Jandel Scientific, San Raphael, CA). Differences in results were considered significant when p ⬍ 0.05.
RESULTS The pilot experiment showed that the intrapleural injection of TGF-2 can produce a pleurodesis (Table 1). All rabbits that received TGF-2 intrapleurally had a pleurodesis score of 4. In contrast, the two rabbits that received one or four injections of the buffer alone had pleurodesis scores of only 0 and 1, respectively. In addition, the pilot study showed that a single injection of TGF-2 at a dose of 5, 10 or 20 g could produce a
TABLE 1 RESULTS FROM THE PILOT STUDY FOR THE INTRAPLEURAL INJECTION OF TRANSFORMING GROWTH FACTOR-2
Rabbit 1 2 3 4 5 6 7 8 9 10
Pleurodesis Score
WBC Count (Cells/mm3)
Volume (ml )
LDH (IU/dl )
Protein (gm/dl )
Dose
No. Injections
Right
Left
24 h
48 h
72 h
Total
24 h
48 h
72 h
24 h
48 h
72 h
24 h
48 h
72 h
Control Control 5 g 5 g 10 g 10 g 20 g 20 g 20 g 20 g
1 4 1 4 1 4 1 2 3 4
0 1 4 4 4 4 4 4 4 4
0 1 0 2 2 2 2 2 3 3
3.5 3 0 24 35 11 24 22 15 17
2 2 7 24 22 30 15 40 28 25
5 5 16 4 10 26 17 10 22 18
18.5 19.3 54.5 81.5 78 120 70 107 131 122
34,362 10,156 — 915 1,467 975 1,376 1,508 1,377 1,147
16,094 4,167 742 350 613 379 492 492 488 506
5,871 7,767 1,013 127 1,346 684 517 661 385 392
6,023 13,601 — 1,799 4,665 7,232 4,620 4,500 7,343 1,342
4,177 4,821 1,622 382 1,654 1,629 1,003 484 783 404
3,496 1,719 3,072 524 877 880 865 853 688 589
3.6 2 — 3.4 3.1 3.1 3.3 2.1 3 3.1
3.6 2.9 3.2 2.8 3.1 2.1 3.2 2.5 3 2.1
3.2 3.3 3.5 2.8 3.3 2.2 3.5 3.1 2.4 2
Definition of abbreviations: LDH ⫽ lactate dehydrogenase; WBC ⫽ white blood cell.
100
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
VOL 162
2000
Figure 3. Mean pleural fluid WBC counts on the first 3 d after intrapleural injection of various amounts of TGF-2 and buffer alone. *p ⬍ 0.05 when compared with control. Figure 1. Degree of pleurodesis after intrapleural injection of buffer and varying doses of TGF-2. Higher doses of TGF-2 were more effective at producing pleurodesis.
pleurodesis. The fibrous reaction with the higher doses was actually greater than would be ideal in producing a pleurodesis. All of the rabbits except the one that received a single dose of 5 g of TGF-2 had some adhesions in the contralateral pleural space. The rabbit that received four separate doses of 20 g of TGF-2 had a fibrin ball in the peritoneal cavity, 112 ml of straw-colored peritoneal fluid, 28 ml of bloody pleural fluid in the right pleural space, and a pleurodesis score of 3 on the left (control) side. Similar findings were made for the rabbit that received three separate doses of 20 g of TGF-2, but this animal had no fibrin balls in the peritoneal cavity. The rabbit that received four separate doses of 10 g TGF-2 had no fibrin balls in the peritoneal cavity, but did have 100 ml of strawcolored peritoneal fluid, 8 ml of pleural fluid, and a pleurodesis score of 2 on the control side. The rabbit that received a single dose of 5 g of TGF-2 had no pleural or peritoneal fluid and had a pleurodesis score of 0 on the control side.
Figure 2. Mean amounts of pleural fluid produced during each of the first 3 d, and the total amount of fluid produced (including at autopsy), after the intrapleural injection of buffer and various amounts of TGF-2. *p ⬍ 0.05 when compared with control; **p ⬍ 0.05 when compared with 5.0 g TGF-2.
In the pilot study, the intrapleural injection of TGF-2 resulted in the formation of large amounts of pleural fluid, which was characterized by a relatively low WBC count and relatively low LDH levels (Table 1). Several of the rabbits had more than 20 ml of pleural fluid present at the 24 h time period. In the primary study, single intrapleural injections of TGF-2 resulted in pleurodesis in a dose-dependent manner (Figure 1). The mean pleurodesis score in the group that received 5 g of TGF-2 (3.6 ⫾ 0.9) was significantly higher than the mean pleurodesis score in the groups that received 1.67 g (2.6 ⫾ 1.4, p ⫽ 0.045), 0.50 g (1.5 ⫾ 1.0, p ⬍ 0.001), of 0.167 g (0.6 ⫾ 0.9, p ⬍ 0.001) of TGF-2, or buffer (0.3 ⫾ 0.1, p ⬍ 0.001). The mean score for the group that received 1.67 g of TGF-2 was significantly greater than that for the group that received 0.50 g (p ⫽ 0.012) or 0.167 g (p ⫽ 0.003) or the buffer solution (p ⬍ 0.001). In none of the rabbits did the mean pleurodesis score on the left (control) side exceed 1. Since the pleurodesis scores were very low and did not differ significantly in the control and 0.167 g TGF-2 groups, these groups were combined for statistical analysis of the pleural fluid results. In the primary study, intrapleural injection of the higher doses of TGF-2 resulted in a high-volume pleural effusion
Figure 4. Mean pleural fluid LDH levels on the first 3 d after intrapleural injection of various amounts of TGF-2 and of buffer alone. *p ⬍ 0.05 when compared with control.
101
Light, Cheng, Lee, et al.: TGF- for Pleurodesis TABLE 2 DIFFERENTIAL CELL COUNTS AT 24, 48, AND 72 h AFTER INJECTION OF BUFFER AND VARYING AMOUNTS OF TRANSFORMING GROWTH FACTOR-2 TGF-2
Neutrophils 24 h 48 h 72 h Monocytes 24 h 48 h 72 h Lymphocytes 24 h 48 h 72 h
Buffer n ⫽ 5
(0.167 g) n ⫽ 4
(0.50 g) n ⫽ 9
(1.67 g) n ⫽ 10
(5.0 g) n ⫽ 12
58.4 ⫾ 4.4 31.3 ⫾ 3.3 28.0 ⫾ 2.1
66.7 ⫾ 5.9 50.0 ⫾ 9.1 27.8 ⫾ 3.3
60.1 ⫾ 2.8 33.4 ⫾ 3.3 27.4 ⫾ 5.4
57.5 ⫾ 4.1 25.0 ⫾ 3.0 34.5 ⫾ 6.4
61.1 ⫾ 6.4 32.2 ⫾ 4.9 25.7 ⫾ 5.5
41.2 ⫾ 4.6 65.3 ⫾ 2.6 68.3 ⫾ 2.3
32.3 ⫾ 6.0 46.8 ⫾ 9.1 68.0 ⫾ 3.3
39.3 ⫾ 2.9 63.7 ⫾ 3.2 68.6 ⫾ 5.3
41.5 ⫾ 4.2 71.7 ⫾ 3.6 62.5 ⫾ 5.9
37.8 ⫾ 6.3 66.7 ⫾ 4.8 71.4 ⫾ 5.6
0.4 ⫾ 0.4 3.3 ⫾ 0.9 3.7 ⫾ 2.7
1.0 ⫾ 0.6 3.3 ⫾ 1.1 4.3 ⫾ 2.7
0.6 ⫾ 0.2 2.8 ⫾ 1.2 4.0 ⫾ 1.3
1.0 ⫾ 0.6 3.3 ⫾ 1.1 3.0 ⫾ 1.6
1.1 ⫾ 0.5 1.1 ⫾ 0.4 2.9 ⫾ 0.7
Definition of abbreviation: TGF-2 ⫽ transforming growth factor-2. Values are given as mean ⫾ SEM.
(Figure 2) with a low WBC (Figure 3). The fluid collected at 24 h was an exudate, in that the mean pleural fluid LDH concentration exceeded 2,000 IU (Figure 4), although the mean protein levels at 24 h were in the 2.7 to 3.0 gm/dl range (data not shown). At 24 h, the amount of pleural fluid was significantly (p ⬍ 0.02) greater in the groups that received the three higher doses of TGF-2 than in the control group. The group that received 5 g of TGF-2 had significantly higher fluid volumes at 48 and 72 h. When the total amount of fluid (including that obtained at autopsy) was compared in the different groups, the mean amount of fluid was significantly greater in the rabbits that received 5 g of TGF-2 than in any of the other groups (p ⬍ 0.05). The increased fluid had decreased to less than 2 ml/d in almost all rabbits by Day 5, and none of the rabbits had pleural fluid present at Day 14. Although the intrapleural injection of TGF-2 induced the accumulation of a large amount of pleural fluid, the fluid did not appear to be particularly inflammatory (Figures 3 and 4). The pleural fluid WBC count and LDH level were both significantly lower in the group that received 5 g of TGF-2 than they were in the control group on all 3 d on which measure-
Figure 5. Mean degree of pleurodesis after intrapleural injection of TGF-2 (5.0 g), doxycycline (10 mg/kg), and talc slurry (400 mg/kg).
ments were made. The differential cell counts were comparable in all groups (Table 2). We next compared the results of the intrapleural injection of 5.0 g of TGF-2 with the results of the injection of doxycycline at 10 mg/kg or talc slurry at 400 mg/kg. Although there were no significant differences in the pleurodesis scores (Figure 5), the pleural fluid that resulted from the intrapleural injection of doxycycline and talc was much smaller in volume (Figure 6) but was more inflammatory than that which resulted from the intrapleural injection of TGF-2 (Figures 7 and 8). At 24 h after injection, the mean pleural fluid WBC counts after the intrapleural injection of TGF-2, talc, and doxycycline were 830 cells/mm3, 6,430 cells/mm3, and 29,149 cells/mm3, respectively. At this same time the mean pleural fluid LDH levels after the intrapleural injection of TGF-2, talc, and doxycycline were 2,965 IU/dl; 26,435 IU/dl, and 29,593 IU/dl, respectively.
DISCUSSION The present study demonstrates that a single intrapleural injection of TGF-2 produces an excellent pleurodesis in rab-
Figure 6. Mean amounts of pleural fluid produced during each of the first 3 d, and the total amount of fluid produced (including at autopsy), after intrapleural injection of TGF-2 (5 g), doxycycline (10 mg/kg), or talc slurry (400 mg/kg). Note that there was much more fluid after the intrapleural injection of TGF-2. *p ⬍ 0.05 when compared with TGF-.
102
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
VOL 162
2000
Figure 7. Mean WBC counts in pleural fluid during each of the first 3 d after intrapleural injection of TGF-2 (5 g), doxycycline (10 mg/kg), or talc slurry (400 mg/kg). *p ⬍ 0.05 when compared with TGF-.
Figure 8. Mean pleural fluid LDH levels during each of the first 3 d after intrapleural injection of TGF-2 (5 g), doxycycline (10 mg/kg), or talc slurry (400 mg/kg). *p ⬍ 0.05 when compared with TGF-.
bits. This pleurodesis is at least as good as that which results from the intrapleural injection of doxycycline at 10 mg/kg or talc slurry at 400 mg/kg. Interestingly, the intrapleural injection of TGF-2 resulted in the production of much more pleural fluid, with a significantly lower LDH concentration and WBC count, than did the injection of doxycycline or talc. These observations suggest that induction of a pleurodesis by the intrapleural injection TGF-2 might produce less injury to the pleura than does the injection of doxycycline or talc slurry, and therefore might produce less chest pain and systemic symptoms. The mechanism by which various agents produce a pleurodesis is largely unknown. The initial event in the production of a pleurodesis is usually an injury to the pleura. An acute exudative pleural effusion develops within 12 h after the instillation of essentially all of the agents that are currently used for inducing pleurodesis, including talc (11), tetracycline derivatives (16), bleomycin (17), and mitoxantrone (18). Subsequently, the resolution of the pleural injury may or may not result in a pleurodesis (17). It is likely that cytokines in the pleural space are responsible for determining whether or not a pleurodesis will result from a pleural injury. We reasoned that if this were true, then the intrapleural injection of a cytokine might produce a pleurodesis without requiring nonspecific injury to the pleura. Among the cytokines, TGF- appears to be a promising candidate for the production of a pleurodesis. TGF- belongs to a superfamily of polypeptide factors that control development and tissue homeostasis in organisms from Drosophila to humans (6). TGF- is found in inflammatory cells (especially macrophages) and platelets, and is very abundant in the cells of the lung. TGF- stimulates ECM accumulation, is chemotactic for fibroblasts and monocytes, and is involved in many lung diseases in which fibrosis plays a part, including idiopathic pulmonary fibrosis (7, 19), asbestotic lung disease (20), and radiation-induced lung disease (21). TGF- has several characteristics which suggest that it might be an ideal agent for inducing pleurodesis. Pleurodesis results from fibrosis, and the excessive production of TGF- can lead to fibrosis. If rats are transfected with an adenovirus vector expressing an active form of TGF-1, the transient overexpression of TGF- results in prolonged and severe interstitial and pleural fibrosis (7). In humans, there are at least five polymorphisms of the TGF-1 gene. The lymphocytes of individuals of different TGF- genotypes produce varying amounts
of TGF-. In a population of lung transplant patients, individuals with the genotypes associated with higher production of TGF- had more fibrosis in their native lung (22). Mesothelial cells express and secrete TGF- (23). Since TGF- induces its own transcription (8), a single intrapleural injection of TGF- could induce mesothelial cells to express and secrete TGF-. Inhibition of the fibrinolytic pathway is thought to be necessary for the production of a pleurodesis (10); the incubation of human pleural mesothelial cells with TGF- results in increased levels of PAI-1 (9), which should inhibit the fibrinolytic system. What is the mechanism by which the intrapleural injection of TGF-2 produces a pleurodesis? Although most agents that produce a pleurodesis induce a pleural injury, we speculate that such a pleural injury is not necessary when TGF-2 is injected intrapleurally to produce a pleurodesis. In our study, the low pleural fluid WBC count and LDH level after injection of TGF-2 suggests that the pleura was not severely injured. Rather, the intrapleural injection of TGF-2 appears to induce mesothelial cells to produce collagen and more TGF-, which in turn leads to the production of additional collagen. Theoretically, it is possible that pleurodesis resulting from the intrapleural injection of TGF-2 could represent an immune response to a non-species–specific product, since we used human TGF-2 in rabbits. We believe that this is unlikely for the following reasons. First, the primary structure of the TGF- isoforms is remarkably conserved among mammalian species (24). For example, there is only one amino acid difference between human and murine TGF-1 in the activated form. Second, there is complete interspecies cross-reactivity of TGF- isoforms with their cognate receptors in all mammals (24). Human TGF-1, -2, and -3 have all shown consistent biologic activity in tissue studies in the mouse, rat, dog, rabbit, and pig. Third, the immunosuppressive aspect of TGF- makes it unlikely that the molecule would evoke a classic immune response in the host. Indeed, it is extraordinarily difficult to raise antibody against this molecule (24). Fourth, when we injected other human proteins (tumor necrosis factor-␣ blocking antibody or interleukin-8) into the pleural space, no pleurodesis resulted. In our study, the intrapleural injection of TGF-2 led to the production of much larger volumes of pleural fluid than did the intrapleural injection of doxycycline or talc. What is the mechanism by which the intrapleural injection of TGF-2 leads to the production of large amounts of pleural fluid? One possibility is that TGF-2 increases the production of pleural
103
Light, Cheng, Lee, et al.: TGF- for Pleurodesis
fluid via its induction of vascular endothelial growth factor (VEGF). When quiescent cultures of mouse embryo-derived cells or human lung adenocarcinoma cells are treated with TGF-, VEGF messenger RNA and protein are induced (25). In addition, when human synovial fibroblasts are incubated with various cytokines, TGF- is the strongest inducer of VEGF secretion (26). VEGF is known to markedly increase the permeability of the vasculature (27). In some studies, VEGF was more potent than histamine in increasing vascular permeability (28). Therefore, if TGF- induced VEGF secretion in the pleural space, the increased levels of VEGF could certainly lead to additional pleural fluid formation. The present study shows that the intrapleural injection of TGF- produces a pleurodesis, but what are the possible side effects? Our pilot study demonstrates that repeated intrapleural injection of large doses of TGF-2 results in some degree of pleurodesis on the contralateral side, and in the accumulation of fibrous balls in the peritoneal cavity. In humans with multiple sclerosis who were given TGF-2 intravenously at up to 2.0 g/kg three times a week for 4 wk, there was a mild deterioration of renal function and mild anemia, which were reversible after the drug was discontinued (29). What are the clinical implications of the present study? Certainly none of the available agents for pleurodesis is ideal. The three primary agents used for pleurodesis at present are bleomycin, talc, and the tetracycline derivatives. The only two U.S. Food and Drug Adminstration-approved agents are bleomycin and talc. Bleomycin is expensive and is relatively ineffective as compared with other sclerosing agents (1). Also, since it does not produce pleurodesis in animals with normal pleurae (17, 30), bleomycin is unlikely to produce a pleurodesis in normal humans. There are significant questions about the safety of talc. In a recent retrospective study of 89 talc pleurodesis procedures in 78 patients, the incidence of respiratory complications or death was 33%; eight patients developed adult respiratory distress syndrome, one patient died, six patients developed dyspnea, and three patients developed reexpansion pulmonary edema (4). The intrapleural injection of a tetracycline derivative is at times very painful, and recent animal studies demonstrate that in rabbits it produces increases in liver enzymes and produces tissue toxicity (31). Accordingly, there is need for an effective nontoxic agent for the production of a pleurodesis. Additional studies in humans are necessary to delineate whether TGF-2 is such an agent. In conclusion, the intrapleural instillation of TGF-2 in rabbits produces an excellent pleurodesis. The ability of TGF- or other cytokines to produce a pleurodesis in humans should be evaluated. If the intrapleural injection TGF- produces a pleurodesis without side effects, the production of a pleurodesis in this manner would certainly be preferable to producing pleurodesis by the injection of a compound such as a tetracycline derivative, which produces a severe chemical burn; the injection of an impure and dangerous chemical such as talc; or the injection of an agent such as bleomycin, which is relatively ineffective and is also expensive. Acknowledgment : The authors would like to acknowledge the editorial assistance of Ms. Sheila Rupp, Dr. Paul Branca, Dr. Marcelo Vaz, Dr. Barrett Conner, and Dr. Dereje-Ayo.
3.
4. 5. 6. 7.
8. 9.
10.
11.
12.
13.
14.
15.
16.
17. 18.
19.
20.
21.
22.
23.
References 1. Walker-Renard, P. B., L. M. Vaughan, and S. A Sahn. 1994. Chemical pleurodesis for malignant pleural effusions. Ann. Intern. Med. 120:56–64. 2. Light, R. W., V. S. O’Hara, T. E. Moritz, A. J. McElhinney, R. Butz, C. M. Haakenson, R. C. Read, C. S. Sassoon, C. E. Eastridge, R. Berger, L. J. Fontenelle, R. H. Bell, S. G. Jenkinson, D. Shure, W. Merrill, E. Hoover, and S. C. Campbell. 1990. Intrapleural tetracycline
24. 25.
for the prevention of recurrent spontaneous pneumothorax. J.A.M.A. 264:2224–2230. Milanez-Campos, J. R., E. C. Werebe, F. S. Vargas, F. B. Jatene, and R.W. Light. 1997. Respiratory failure due to insufflated talc. Lancet 349:251–252. Rehse, D. H., R. W. Aye, and M. G. Florence. 1999. Respiratory failure following talc pleurodesis. Am. J. Surg. 177:437–440. Light, R. W., and F. S. Vargas. 1997. Pleural sclerosis for the treatment of pneumothorax and pleural effusion. Lung 175:213–223. Grande, J. P. 1997. Role of transforming growth factor- in tissue injury and repair. Proc. Soc. Exp. Biol. Med. 214:27–40. Sime, P. J., Z. Xing, F. L. Graham, K. G. Csaky, and J. Gauldie. 1997. Adenovector-mediated gene transfer of active transforming growth factor-beta1 induces prolonged severe fibrosis in rat lung. J. Clin. Invest. 100:768–776. Perkett, E. A. 1995. Role of growth factors in lung repair and diseases. Curr. Opin. Pediatr. 7:242–249. Idell, S., C. Zwieb, A. Kumar, K. B. Koenig, and A. R. Johnson. 1992. Pathways of fibrin turnover of human pleural mesothelial cells in vitro. Am. J. Respir. Cell Mol. Biol. 7:414–426. Rodriguez-Panadero, F., A. Segado, J. Martin Juan, R. Ayerbe, I. Torres Garcia, and J. Castillo. 1995. Failure of talc pleurodesis is associated with increased pleural fibrinolysis. Am. J. Respir. Crit. Care Med. 151: 785–790. Xie, C., L. R. Teixeira, N.-S. Wang, J. P. McGovern, and R. W. Light. 1998. Serial observations after high dose talc slurry in the rabbit model for pleurodesis. Lung 176:299–307. Light, R. W., N. S. Wang, C. S. H. Sassoon, S. E. Gruer, and F. S. Vargas. 1994. Comparision of the effectiveness of tetracycline and minocycline as pleural sclerosing agents in rabbits. Chest 106:577–582. Xie, C., L. R. Teixeira, J. P. McGovern, and R. W. Light. 1998. Systemic corticosteroids decrease the effectiveness of talc pleurodesis. Am. J. Respir. Crit. Care Med. 157:1441–1444. Ogawa, Y., and S. M. Seyedin. 1991. Purification of transforming growth factors beta 1 and beta 2 from bovine bone and cell culture assays. Methods Enzymol. 198:317–327. Rogers, J. T., D. Cheng, A. Wheeler, L. Teixeira, and R. W. Light. 1998. The effects of tumor necrosis factor alpha (TNF␣) blocking antibody on pleurodesis in rabbits. Chest 114:260S. Wu, W., L. R. Teixeira, and R. W. Light. 1998. Doxycycline pleurodesis in rabbits: comparison of results with and without chest tube. Chest 114:563–568. Sahn, S. A., and J. T. Good. 1981. The effect of common sclerosing agents on the rabbit pleural space. Am. Rev. Respir. Dis. 124:65–67. Vargas, F. S., L. R. Teixeira, L. Antonangelo, L. M. M. F. Silva, C. M. C. Strunz, and R. W. Light. 1998. Acute and chronic pleural changes after the intrapleural instillation of mitoxantrone in rabbits. Lung 176: 227–236. Coker, R. K., G. J. Laurent, S. Shahzeidi, P. A. Lympany, R. M. du Bois, P. K. Jeffery, and R. J. McAnulty. 1997. Transforming growth factorsbeta 1, -beta 2, and -beta 3 stimulate fibroblast procollagen production in vitro but are differentially expressed during bleomycin-induced lung fibrosis. Am. J. Pathol. 150:981–991. Jagirdar, J., T. C. Lee, J. Reibman, L. I. Gold, C. Aston, R. Begin, and W. N. Rom. 1997. Immunohistochemical localization of transforming growth factor beta isoforms in asbestos-related diseases. Environ. Health Perspect. 105(Suppl. 5):1197–1203. Anscher, M. S., F. M. Kong, K. Andrews, R. Clough, L. B. Marks, G. Bentel, and R. L. Jirtle. 1998. Plasma transforming growth factor beta1 as a predictor of radiation pneumonitis. Int. J. Radiat. Oncol. Biol. Phys. 41:1029–1035. Awad, M. R., A. El-Gamel, P. Hasleton, D. M. Turner, P. J. Sinnott, and I. V. Hutchinson. 1998. Genotypic variation in the transforming growth factor-beta1gene: association with transforming growth factor-beta1 production, fibrotic lung disease and graft fibrosis after lung transplantation. Transplantation 66:1014–1020. Gerwin, B. I., J. F. Lechner, R. R. Reddel, A. B. Roberts, K. C. Robbins, E. W. Gabrielson, and C. C. Harris. 1987. Comparison of production of transforming growth factor-beta and platelet-derived growth factor by normal human mesothelial cells and mesothelioma cell lines. Cancer Res. 47:6180–6184. Roberts, A. B. 1998. Molecular and cell biology of TGF-beta. Miner. Electrolyte Metab. 24:111–119. Pertovaara, L., A. Kaipainen, T. Mustonen, A. Orpana, N. Ferrara, O. Saksela, and K. Alitalo. 1994. Vascular endothelial growth factor is induced in response to transforming growth factor-beta in fibroblastic and epithelial cells. J. Biol. Chem. 269:6271–6274.
104
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
26. Berse, B., J. A. Hunt, R. J. Diegel, P. Morganelli, K. Yeo, F. Brown, and R. A. Fava. 1999. Hypoxia augments cytokine transforming growth factor-beta (TGF-beta and IL-1)-induced vascular endothelial growth factor secretion by human synovial fibroblasts. Clin. Exp. Immunol. 115:176–182. 27. Brown, L. F., M. Detmar, K. Claffey, and J. A. Nagy. 1997. Vascular permeability factor/vascular endothelial growth factors: a multifunctional angiogenic cytokine. EXS 79:233–269. 28. Collins, P. O., D. T. Connolly, and T. J. Williams. 1993. Characterization of increase in vascular permeability induced by vascular permeability factor in vivo. Br. J. Pharmacol. 109:195–199.
VOL 162
2000
29. Calabresi, P. A., N. S. Fields, H. W. Maloni, A. Hanham, J. Carlino, J. Moore, M. C. Levin, S. Dhib-Jalbut, L. R. Tranquill, H. Austin, H. F. McFarland, and M. K. Racke. 1998. Phase I trial of transforming growth factor beta-2 in chronic progressive MS. Neurology 51:289–292. 30. Vargas, F. S., N.-S. Wang, H. M. Lee, S. E. Gruer, C. S. H. Sassoon, and R. W. Light. 1993. Effectiveness of bleomycin in comparison to tetracycline as pleural sclerosing agent in rabbits. Chest 104:1582–1584. 31. Mitchem, R. E., B. L. Herndon, R. M. Fiorella, A. Molteni, C. N. Battie, and G. R. Reisz. 1999. Pleurodesis by autologous blood, doxycycline, and talc in a rabbit model. Ann. Thorac. Surg. 67:917–921.