American Journal of Respiratory and Critical Care Medicine

Pneumonectomized rats develop pulmonary hypertension (PH) and pulmonary vascular neointimal formation 4 wk after monocrotaline (MCT) administration. Male Sprague-Dawley rats were injected with MCT (60 mg/kg) on Day 7 after left pneumonectomy. Three groups (n = 5) received 40-O-(2-hydroxyethyl)-rapamycin (RAD, 2.5 mg/kg/d, by gavage): Group PMR5–35 from Day 5 to Day 35, Group PMR5–14 from Day 5 to Day 14, and Group PMR15–35 from Day 15 to Day 35. By Day 35, rats that received vehicle had higher mean pulmonary arterial pressures (Ppa = 41 ± 3 mm Hg) (p < 0.001), right ventricular systolic pressures (Prv,s = 45 ± 2 mm Hg) (p < 0.01), and right ventricle/(left ventricle plus septum) (0.55 ± 0.05) (p = 0.028) than rats in Groups PMR5–35 (Ppa = 25 ± 3 mm Hg, Prv,s = 32 ± 7 mm Hg, RV/LV&S = 0.42 ± 0.06) and PMR5–14 (Ppa = 29 ± 4 mm Hg, Prv,s = 30 ± 5 mm Hg, RV/LV&S = 0.43 ± 0.07). Pulmonary arterial neointimal formation (quantified by a vascular occlusion score) was more severe in vehicle-treated rats (1.93 ± 0.03) than in Groups PMR5–14 (1.56 ± 0.27) and PMR5–35 (1.57 ± 0.1) (p < 0.01). RAD attenuates the development of MCT-induced pulmonary arterial hypertension in the pneumonectomized rat.

Primary pulmonary hypertension (PPH) is a devastating and incurable disease that leads to relentless elevations in pulmonary arterial pressures and death due to circulatory failure (1, 2). The pathology of PPH is characterized by abnormal expansions of endothelial cells, medial hypertrophy, and adventitial thickening of pulmonary arteries (3). Each appears to play a role in the development and progression of pulmonary hypertension (PH). (3). In spite of research that suggests an important role for endothelial proliferation and neointimal formation in the development and progression of PH, there are no therapeutic strategies designed to suppress neointimal formation being used to treat pulmonary hypertension (4-6). The current medical management of PPH is directed at anticoagulation and vasodilatation rather than the prevention of endothelial proliferation and neointimal formation (2). Prostacyclin may have beneficial effects on vascular remodeling, because some patients who do not demonstrate a vasodilator response to prostacyclin appear to benefit from its use (2, 5, 7).

Monocrotaline (MCT) is a toxin derived from plants of the Crotalaria species (8). When MCT is injected into rats it causes pulmonary arterial endothelial cell injury, pulmonary artery medial hypertrophy, and PH (8, 9). The average pulmonary arterial pressure (Ppa) rises to approximately 32 mm Hg, 4 wk after monocrotaline administration (10, 11). The combination of monocrotaline administration with pneumonectomy results in higher Ppa (mean Ppa, 45 mm Hg) in addition to pulmonary arterial neointimal formation (12). The development of pulmonary vascular neointimal formation in this rat model of pulmonary hypertension is thought to be due to increases in shear stresses (the entire cardiac output flows to the right lung) because neointimal formation does not occur after monocrotaline toxin alone (12). Disturbed laminar shear stresses on cultured endothelial cells have been shown to activate transcription factors associated with inflammation and cell proliferation, including nuclear factor κB (NF-κB), early growth response protein 1 (Egr-1), and activator protein 1 (AP-1, composed of Fos and Jun proteins) (13).

Rapamycin is a macrolide immunosuppressant that is currently being used as a novel therapy for chronic allograft rejection (14). Rapamycin exerts antiproliferative effects on lymphoid and nonlymphoid cells by inhibiting growth factor receptor-mediated signaling at the level of protein translation. The target of rapamycin (TOR) is large protein kinase of the phosphoinositide 3-kinase family (15-17). Rapamycin binding to TOR interferes with downstream activation of ribosomal p70 S6 kinase, and with phosphorylation and activation of translation initiation factor 4E-binding protein 1 (4E-BP1) (15-17). Rapamycin has been shown to prevent vascular endothelial growth factor (VEGF)-mediated and serum-stimulated proliferation of endothelial cells in vitro (18, 19). In addition, rapamycin reduced the arterial proliferative response, intimal thickening, and vascular smooth muscle proliferation after angioplasty in pigs (20, 21). We hypothesized that rapamycin might prevent pulmonary vascular neointimal formation associated with shear stress and thereby attenuate the development of pulmonary hypertension. This study investigates the effects of 40-O-(2-hydroxyethyl)-rapamycin (RAD), an orally active derivative of rapamycin (22), on monocrotaline-induced PH in the pneumonectomized rat.

Study Sample

Twenty pathogen-free, 13-wk-old, male Sprague-Dawley rats (body weight, 350–400 g) were studied.

Left Pneumonectomy

On Day 0, rats were anesthesized with atropine sulfate (50 μg, intramuscular), ketamine hydrochloride (10 mg, intramuscular), and xylazine (3 mg, subcutaneous). After oral endotracheal intubation (with a 14-gauge catheter; Baxter, Deerfield, IL), anesthesia was maintained with halothane inhalation (0.5%) and rats were ventilated with a Harvard rodent ventilator (tidal volume, 3.0 ml; respiratory rate, 60/min; positive end-expiratory pressure [PEEP], 1 cm H2O) (Type 683; Harvard Apparatus, South Natick, MA). Left pneumonectomy was performed via left thoracotomy, using aseptic technique.

Monocrotaline Administration

On Day 7, rats were injected subcutaneously in the right hindlimb with monocrotaline (MCT, 60 mg/kg; Sigma, St. Louis, MO) (dissolved in distilled water, adjusted to pH 7.40 with 0.5 N HCl).

Treatment Groups

Rats were randomized to receive 40-O-(2-hydroxyethyl)-rapamycin (RAD; gift of Novartis Pharma, Basel, Switzerland), or vehicle by daily gavage. Five groups were studied: Group PMR5–35 received RAD (2% solution, 2.5 mg/kg/d) from Day 5 to Day 35 (n = 5), Group PMR5–14 (early treatment group) received RAD (2.5 mg/kg/d) from Day 5 to Day 14 (n = 5), and Group PMR15–35 (late treatment group) received RAD (2.5 mg/kg/d) from Day 15 to Day 35 (n = 5). Group PMV (vehicle group) received a vehicle (placebo) solution from Day 5 to Day 35 (n = 5). An additional five rats were studied as a control (N) group. These rats did not undergo pneumonectomy, nor did they receive monocrotaline or RAD.

All animals received humane care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research, and the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Science and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1985). The study was approved by the Stanford Panel on Laboratory Animal Care and was conducted in compliance with Stanford University regulations.

Hemodynamic Studies and Tissue Preparation

On Day 35, rats were anesthesized by intramuscular injections of atropine sulfate (50 μg, intramuscular) and ketamine hydrochloride (10 mg, intramuscular) and placed in the supine position. Anesthesia was maintained with inhalation halothane (0.5%, by hood). A carotid arterial catheter (PE-50, 0.58-mm i.d.) was placed after cutdown. A pulmonary artery catheter (PV 1, 0.28-mm i.d.) was inserted into the right internal jugular vein through an introducer. The pulmonary artery catheter was passed (under pressure wave guidance) through the right ventricle into the pulmonary artery. Right ventricular systolic pressure measurements were also obtained by percutaneous needle (27 gauge) puncture of the right ventricle. Mean arterial blood pressure, pulmonary artery blood pressure, and right ventricular systolic blood pressure were recorded. After exsanguination, the right lung, right ventricle, left ventricle plus septum, liver, spleen, kidney, testis, and thymus were collected for histology. Tissues were fixed in 10% neutral buffered formalin. The lungs were axially sectioned, processed, and embedded in paraffin wax. Five-micron sections were prepared and stained with elastin-van Gieson (EVG). The severity of neointimal formation was scored according to the criteria of Okada and coworkers (12). Briefly, the absence of neointimal formation equals 0; the presence of neointimal proliferation causing less than 50% lumenal narrowing equals 1; lumenal narrowing greater than 50% equals 2. To facilitate comparisons across groups of rats, organ weights are presented per kilogram of body weight.

Statistical Analysis

Data are presented as means ± standard deviation. First, the data from normal rats were compared with data for Group PMV (the disease model), using the Student t test (statistical significance was indicated by p < 0.05). Next, Groups PMV, PMR5–35, PMR5–14, and PMR15–35 were analyzed by two-way analysis of variance (ANOVA) and multiple comparisons in order to determine the effects of early and late therapy. A value of p < 0.05 was considered significant.

Twenty-five rats were studied. Five rats were used as a control group. Twenty rats received monocrotaline (60 mg/kg, subcutaneous) 7 d after left pneumonectomy. One rat in Group PMV and one rat in Group PMR15–35 died from mediastinitis (secondary to gavage tube trauma) on postoperative Days 14 and 21, respectively, and were excluded from the analysis. The remaining 18 rats underwent hemodynamic measurements and were killed on postoperative Day 35. The four treatment groups did not differ in terms of body mass.

Hemodynamic Parameters (Day 35)

Pulmonary arterial pressure. Rats that received RAD had lower mean Ppa than rats that received vehicle (Figure 1A). Rats that received RAD from Day 5 to Day 35 (n = 5) had the lowest Ppa (25 ± 3 mm Hg). Rats that received RAD from Day 5 to Day 14 (n = 5) had lower Ppa (29 ± 4 mm Hg) than those treated from Day 15 to Day 35 (n = 4) (38 ± 5 mm Hg). Of the four treatment groups, vehicle-treated rats (n = 4) had the highest Ppa (41 ± 3 mm Hg). Two-way ANOVA and multiple comparison revealed that the main effect of therapy occurred between Days 5 and 14 (Group PMR5–14) (p < 0.001).

Right ventricular systolic pressure. Rats that received RAD had lower right ventricular systolic pressure (Prv,s) than rats that received vehicle (Figure 1B). Rats that received RAD from Day 5 to Day 35 had a Prv,s (32 ± 7 mm Hg). Rats that received RAD from Day 5 to Day 14 had lower Prv,s (30 ± 5 mm Hg) than those treated from Day 15 to Day 35 (42 ± 7 mm Hg). Of the four treatment groups, vehicle-treated rats had the highest Prv,s (45 ± 2 mm Hg). Two-way ANOVA and multiple comparison revealed that the main effect of therapy occurred between Days 5 and 14 (Group PMR5–14) (p < 0.01).

Mean arterial blood pressure measurements were not significantly different between the four groups. Note that values for Group N (a group of five normal, healthy rats) were included as a reference control. Group N demonstrated systemic blood pressures (126 ± 9 mm Hg), pulmonary arterial blood pressures (17 ± 1 mm Hg), and right ventricular systolic blood pressures (25 ± 2 mm Hg) that were typical of healthy adult rats (23, 24) (Figure 1). Measurements of blood hematocrit were not significantly different among the four groups.

Organ Weights

Right ventricular hypertrophy. The development of chronic pulmonary arterial hypertension results in a compensatory hypertrophy of the right ventricle (increased ratio of [right ventricle]/[left ventricle and septum] [RV/LV&S]). Rats that received RAD had lower (RV/LV&S) than did rats that received vehicle (Group PMV = 0.55 ± 0.05) (Figure 2). The mean ratios for the RAD-treated animals were as follows: Group PMR5–35, 0.42 ± 0.06; Group PMR5–14, 0.43 ± 0.07; and Group PMR15–35, 0.45 ± 0.08. Two-way ANOVA and multiple comparison revealed that significant effects of therapy occurred between Days 5 and 14 (Group PMR5–14) (p < 0.01) and between Days 15 and 35 (Group PMR15–35) (p = 0.038).

Other organs. Rats that received RAD had lower thymus weights than rats that received vehicle (Figure 3A). Two-way ANOVA and multiple comparison revealed that the main effect of therapy occurred between Days 15 and 35 (p = 0.012). Rats in Group PMR5–35 had lower spleen weights than rats in Group PMV (Figure 3B). Two-way ANOVA and multiple comparison revealed there was a significant interaction between early and late treatments (p = 0.04). Liver, kidney, and testis weights were not significantly different among treatment groups.

Histopathology

A quantitative analysis of lumenal obstruction on 50 consecutive small pulmonary arteries from each rat in Groups PMV and PMR was performed (Figure 4). The distribution of the vascular lesions, and an average vascular occlusion score (between 0 and 2), are presented (Figure 4). Two-way ANOVA and multiple comparison revealed that significant effects of therapy occurred between Days 5 and 14 (Group PMR5–14) (p < 0.01).

In summary, these data demonstrate that RAD therapy, when administered at the same time as monocrotaline, results in significantly lower pulmonary arterial pressures, less right ventricular hypertrophy, and a lower vascular occlusion score in this rat model. When initiated more than 1 wk after monocrotaline administration (Group PMR15–35), RAD treatment decreases right ventricular hypertrophy.

In this investigation, we demonstrate that treatment with an oral derivative of rapamycin (RAD) attenuates the development of pulmonary hypertension, right ventricular hypertrophy, and pulmonary vascular neointimal formation in pneumonectomized rats that receive monocrotaline. Moreover, we demonstrate that early therapy with RAD (simultaneous with the administration of monocrotaline) appears more effective at preventing vascular remodeling and PH than late therapy (commencing 1 wk after monocrotaline administration). Our data suggest that RAD therapy is most effective when initiated before monocrotaline administration. Late therapy (beginning 1 wk after monocrotaline) does not significantly attenuate the development of pulmonary arterial neointimal formation and pulmonary hypertension but does attenuate the development of right ventricular hypertrophy. Our laboratory has previously demonstrated that right ventricular hypertrophy (RVH) is present as early as Day 21 (RV/LV&S = 0.4) in this disease model (25). Therefore, the effect of late RAD treatment probably results from direct effects on heart muscle proliferation rather than from changes in pulmonary vascular resistance.

The mechanisms that result in neointimal formation in this disease model are still unknown. In rats, monocrotaline administration leads to vascular smooth muscle cell hypertrophy and an increase in medial wall thickness (9, 26). It is believed that an increase in vascular shear stress (left pneumonectomy results in a relative increase in blood flow to the remaining lung) is necessary to produce neointimal formation, because the lesion does not occur after monocrotaline administration alone (12). This pathologic neointimal formation leads to lumenal occlusion and marked increases in pulmonary vascular resistance. Okada and coworkers (27) demonstrated that the angiotensin-converting enzyme inhibitor, quinapril, effectively delayed the development of PH and RVH in a rat model that involved monocrotaline injection followed by pneumonectomy. Five weeks after monocrotaline administration, pneumonectomized rats showed Ppa = 45 ± 5 mm Hg, RV/LV&S = 0.6 ± 0.1, and vascular occlusion score = 1.67 ± 0.08 (27). These values closely match our results in Group PMV, measured 4 wk after monocrotaline administration. Early treatment with quinapril (30 mg/kg/d, commencing 10 d before monocrotaline) resulted in Ppa = 21 ± 3 mm Hg, RV/LV&S = 0.39 ± 0.05, and vascular occlusion score = 0.27 ± 0.09 (27). Delayed treatment with quinapril (initiated 3 wk after MCT) resulted in Ppa = 28 ± 3 mm Hg, RV/LV&S = 0.48 ± 0.1, and vascular occlusion score = 1.20 ± 0.26 (27). The current data indicate that RAD is as effective as quinapril in this disease model when therapy is commenced before monocrotaline administration. However, RAD is probably less effective than quinapril when treatment is started after monocrotaline administration.

There is no evidence to suggest that immunosuppression has a beneficial effect on the development of monocrotaline-induced PH. Several investigations have suggested that T lymphocytes probably do not contribute to the development of monocrotaline-induced PH. First, monocrotaline-induced pulmonary hypertension has been reported to be more severe, rather than less severe, in athymic rats, compared with euthymic littermates, suggesting that T lymphocytes are not required to develop the disease (28). Second, neither the adoptive transfer of lymphocytes, nor anti-lymphocyte serum, has a significant effect on the development or progression of monocrotaline-induced pulmonary hypertension (29). In addition, therapy with the immunosuppressant cyclosporin does not protect against the development of monocrotaline-induced PH (29). Unlike cyclosporin, rapamycin inhibits growth factor receptor-mediated proliferation in both hematopoietic and nonhematopoietic cells (including endothelial cells) (18, 19). There are no data on the impact of the immune system on the development and progression of neointimal lesions in pneumonectomized rats that receive monocrotaline. In other disease models it appears that rapamycin and RAD have effects on vascular remodeling that are independent of immunosuppression (30, 31). Chronic graft vascular disease in rat cardiac allografts is generally unresponsive to immunosuppression, but it is reversed by rapamycin (14). Rapamycin also inhibits the arterial proliferative response after balloon angioplasty in pigs, a process that is also generally unresponsive to immunosuppression (20, 21). On the basis of the above, we believe that the efficacy of RAD in this disease model is likely due to its direct antiangiogenic effects, rather than to immunosuppressive effects.

Current therapies for PPH include vasodilators (such as calcium channel blockers and prostacyclin) and anticoagulation (2). Currently, there are no data on the use of nonvasoactive antiproliferative agents to treat PPH. Rapamycin and RAD have not been used as therapy for PH in patients. In this study, RAD helps to prevent (but does not appear to reverse) vascular neointimal formation in this model of pulmonary hypertension. Most cases of human pulmonary hypertension are advanced by the time of diagnosis. Therefore, rather than curing pulmonary hypertension, RAD might help to prevent further disease progression. This study, and our study of the effect of triptolide in this disease model (25), demonstrate the usefulness of antiangiogenic compounds in the development and progression of pulmonary hypertension, at least in rats. Such strategies might represent a new approach to the management of pulmonary hypertension in humans.

The authors thank Professor Yoshinori Fujii (Department of Statistics, Stanford University) for statistical advice and Gail V. Benson for technical assistance, and the members of the Pearl and Kao laboratories for helpful discussions.

Supported by a gift from the Donald E. and Delia B. Baxter Foundation and by NIH grants AI39624 and HL62588 to P.N.K.

1. Lilienfeld DE, Rubin LJMortality from primary pulmonary hypertension in the United States, 1979–1996. Chest1172000796800
2. Rubin LJPrimary pulmonary hypertension. N Engl J Med3361997111117
3. Higenbottam T, Naeije R, Voelkel NF, Botney MD, Christman B, Giald A, Hales CA, Herve P, Loscalzo J, Weir EK. Pathobiology of pulmonary hypertension. In: Rich S, editor. Primary pulmonary hypertension: executive summary from the World Symposium on Primary Pulmonary Hypertension 1998. Available from the World Health Organization at http://www.who.int/ncd/cvd/pph.html
4. Gurubhagavatula I, Palevsky HIPulmonary hypertension in systemic autoimmune disease. Rheum Dis Clin North Am231997365394
5. Voelkel NF, Tuder RMSevere pulmonary hypertensive diseases: a perspective (in process citation). Eur Respir J14199912461250
6. Cool CD, Stewart JS, Werahera P, Miller GJ, Williams RL, Voelkel NF, Tuder RMThree-dimensional reconstruction of pulmonary arteries in plexiform pulmonary hypertension using cell-specific markers: evidence for a dynamic and heterogeneous process of pulmonary endothelial cell growth. Am J Pathol1551999411419
7. Higenbottam T, Butt AY, McMahon A, Westerbeck R, Sharples LLong-term intravenous prostaglandin (epoprostenol or iloprost) for treatment of severe pulmonary hypertension. Heart801998151155
8. Rosenberg HC, Rabinovitch MEndothelial injury and vascular reactivity in monocrotaline pulmonary hypertension. Am J Physiol2551988H1484H1491
9. van Suylen RJ, Smits JF, Daemen MJPulmonary artery remodeling differs in hypoxia- and monocrotaline-induced pulmonary hypertension. Am J Respir Crit Care Med157199814231428
10. Hill LL, Pearl RGCombined inhaled nitric oxide and inhaled prostacyclin during experimental chronic pulmonary hypertension. J Appl Physiol86199911601164
11. Mathew R, Zeballos GA, Tun H, Gewitz MHRole of nitric oxide and endothelin-1 in monocrotaline-induced pulmonary hypertension in rats. Cardiovasc Res301995739746
12. Okada K, Tanaka Y, Bernstein M, Zhang W, Patterson GA, Botney MDPulmonary hemodynamics modify the rat pulmonary artery response to injury: a neointimal model of pulmonary hypertension. Am J Pathol151199710191025
13. Nagel T, Resnick N, Dewey CF, Gimbrone MAVascular endothelial cells respond to spatial gradients in fluid shear stress by enhanced activation of transcription factors. Arterioscler Thromb Vasc Biol19199918251834
14. Poston RS, Billingham M, Hoyt EG, Pollard J, Shorthouse R, Morris RE, Robbins RCRapamycin reverses chronic graft vascular disease in a novel cardiac allograft model. Circulation10019996774
15. Abraham RTMammalian target of rapamycin: immunosuppressive drugs uncover a novel pathway of cytokine receptor signaling. Curr Opin Immunol101998330336
16. Dennis PB, Fumagalli S, Thomas GTarget of rapamycin (TOR): balancing the opposing forces of protein synthesis and degradation. Curr Opin Genet Dev919994954
17. Kuruvilla FG, Schreiber SLThe PIK-related kinases intercept conventional signaling pathways. Chem Biol61999R129R136
18. Yu Y, Sato JDMAP kinases, phosphatidylinositol 3-kinase, and p70 S6 kinase mediate the mitogenic response of human endothelial cells to vascular endothelial growth factor. J Cell Physiol1781999235246
19. Vinals F, Chambard JC, Pouyssegur Jp70 S6 kinase-mediated protein synthesis is a critical step for vascular endothelial cell proliferation. J Biol Chem27419992677626782
20. Gallo R, Padurean A, Jayaraman T, Marx S, Roque M, Adelman S, Chesebro J, Fallon J, Fuster V, Marks A, Badimon JJInhibition of intimal thickening after balloon angioplasty in porcine coronary arteries by targeting regulators of the cell cycle. Circulation99199921642170
21. Burke SE, Lubbers NL, Chen YW, Hsieh GC, Mollison KW, Luly JR, Wegner CDNeointimal formation after balloon-induced vascular injury in Yucatan minipigs is reduced by oral rapamycin. J Cardiovasc Pharmacol331999829835
22. Schuler W, Sedrani R, Cottens S, Haberlin B, Schulz M, Schuurman HJ, Zenke G, Zerwes HG, Schreier MHSDZ RAD, a new rapamycin derivative: pharmacological properties in vitro and in vivo (see comments). Transplantation6419973642
23. Kay JMEffect of intermittent normoxia on chronic hypoxic pulmonary hypertension, right ventricular hypertrophy, and polycythemia in rats. Am Rev Respir Dis12119809931001
24. Pan LC, Wilson DW, Segall HJStrain differences in the response of Fischer 344 and Sprague-Dawley rats to monocrotaline induced pulmonary vascular disease. Toxicology7919932135
25. Faul JL, Nishimura T, Berry GJ, Benson GV, Pearl RG, Kao PNTriptolide attenuates pulmonary arterial hypertension and neointimal formation in rats. Am J Respir Crit Care Med162200022522258
26. Rabinovitch MInvestigational approaches to pulmonary hypertension. Toxicol Pathol191991458469
27. Okada K, Bernstein ML, Zhang W, Schuster DP, Botney MDAngiotensin-converting enzyme inhibition delays pulmonary vascular neointimal formation. Am J Respir Crit Care Med1581998939950
28. Miyata M, Sakuma F, Ito M, Ohira H, Sato Y, Kasukawa RAthymic nude rats develop severe pulmonary hypertension following monocrotaline administration. Int Arch Allergy Immunol1212000246252
29. Bruner LH, Bull RW, Roth RAThe effect of immunosuppressants and adoptive transfer in monocrotaline pyrrole pneumotoxicity. Toxicol Appl Pharmacol911987112
30. Cole OJ, Shehata M, Rigg KMEffect of SDZ RAD on transplant arteriosclerosis in the rat aortic model. Transplant Proc30199822002203
31. Mohacsi PJ, Tuller D, Hulliger B, Wijngaard PLDifferent inhibitory effects of immunosuppressive drugs on human and rat aortic smooth muscle and endothelial cell proliferation stimulated by platelet-derived growth factor or endothelial cell growth factor (see comments). J Heart Lung Transplant161997484492
Correspondence and requests for reprints should be addressed to Peter N. Kao, M.D., Ph.D., Pulmonary and Critical Care Medicine, Stanford University Medical Center, Stanford, CA 94305-5236. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
163
2

Click to see any corrections or updates and to confirm this is the authentic version of record