Ian S. Paterson

Lower torso ischemia-induced lung injury is leukocyte dependent.

Lower torso ischemia leads on reperfusion to sequestration of polymorphonuclear leukocytes (PMN) in the lungs and increased permeability. This study tests the role of circulating leukocytes (WBC) in mediating this lung injury. Anesthetized sheep prepared with chronic lung lymph fistulae underwent 2 hours of bilateral hind limb tourniquet ischemia. In untreated controls (n = 7), 1 minute after reperfusion there were transient increases in mean pulmonary arterial pressure (MPAP) from 13 to 38 mmHg (p less than 0.05) and pulmonary microvascular pressure (Pmv) from 7 to 18 mmHg (p less than 0.05), changes temporally related to a rise in plasma thromboxane (Tx) B2 levels from 211 to 735 pg/ml (p less than 0.05). Lung lymph TxB2 levels rose from 400 to 1005 pg/ml at 30 minutes (p less than 0.05), and remained elevated longer than plasma levels. Lung lymph flow (QL) rose from 4.3 to 8.3 ml/30 minutes (p less than 0.05) after 30 minutes of reperfusion and remained elevated for 2 hours. The lymph/plasma (L/P) protein ratio was unchanged from 0.6, while the lymph protein clearance increased from 2.6 to 4.6 ml/30 minutes (p less than 0.05), suggesting increased microvascular permeability. WBC counts decreased within the first hour of reperfusion from 6853 to 3796/mm3 (p less than 0.05), and lung histology after 2 hours showed proteinaceous exudates and leukosequestration of 62 PMN/10 high-powered fields (HPF), higher than the 22 PMN/10 HPF (p less than 0.05) in sham animals (n = 3). Recruitment of the pulmonary vasculature by left atrial balloon inflation (n = 3) resulted in a rise in MPAP to 20 mmHg. After 3 hours of balloon inflation, QL stabilized at 9.8 ml/15 minutes, and a pressure-independent L/P protein ratio of 0.3 was achieved. During reperfusion, QL increased further to 11.2 ml/15 minutes, the L/P ratio rose to 0.56 and the calculated osmotic reflection coefficient decreased from 0.70 to 0.44, documenting an increase in lung microvascular permeability. In contrast to these untreated ischemic controls, sheep (n = 7) rendered leukopenic with hydroxyurea or nitrogen mustard and having a total WBC count of 760/mm3 and PMN count of 150/mm3 did not manifest reperfusion-induced increases in MPAP, Pmv, QL, lymph protein clearance, or lung lymph. TxB2 level (p less than 0.05). Plasma TxB2 levels rose slightly at 30 minutes from 199 to 288 pg/ml (p less than 0.05). Lung histology was normal. These data indicate that WBC mediate the ischemia-induced increase in pulmonary microvascular permeability.

Noncardiogenic pulmonary edema after abdominal aortic aneurysm surgery.

Limb ischemia in experimental animals leads to white blood cell (WBC) and thromboxane (Tx)A2 dependent pulmonary dysfunction. This study examines the pulmonary sequelae of lower torso ischemia in 20 consecutive patients aged 63 +/- 5 years (mean +/- SEM) who underwent elective abdominal aortic aneurysm surgery. After 30 minutes of aortic cross-clamping, plasma TxB2 levels had risen from 77 +/- 26 pg/ml to 359 +/- 165 pg/ml (p less than 0.01) and was temporally related to increases in mean pulmonary artery pressure (MPAP) from 18 +/- 1 to 23 +/- 3 mmHg (p less than 0.01), as well as to increases in pulmonary vascular resistance (PVR) from 0.07 +/- 0.02 to 0.12 +/- 0.02 mmHg sec/ml (p less than 0.01). Each time that the aortic clamp was repositioned and with final declamping, after 83 +/- 10 minutes, there were further increases in MPAP to a peak of 32 +/- 2 mmHg (p less than 0.01) and in PVR to 0.26 +/- 0.030 mmHg sec/ml (p less than 0.01), corresponding to a plasma TxB2 level of 406 +/- 177 pg/ml (p less than 0.01). MPAP and PVR returned to baseline values within 30 minutes of declamping. Ten minutes after removal of the aortic clamp, platelet levels had fallen from 180 +/- 41 to 97 +/- 17 X 10(3)/mm3 (p less than 0.01) and WBC levels from 8900 +/- 1100 to 4700 +/- 400/mm3 (p less than 0.01). Both platelets and WBC returned towards normal levels, but at 24 hours, while WBC was elevated at 13000 +/- 900/mm3 (p less than 0.01), platelets were 44% of baseline at 135 +/- 14 X 10(3)/mm3 (p less than 0.01). Four to 8 hours after surgery, pulmonary dysfunction was manifest by increases in physiologic shunt from 9 +/- 2% to 16 +/- 2% (p less than 0.01), and peak inspiratory pressure (PIP) from 23 +/- 2 to 33 +/- 2 cmH2O (p less than 0.01). Chest radiography demonstrated interstitial pulmonary edema in all patients, whereas pulmonary artery wedge pressure was 12 +/- 2 mmHg, excluding the possibility of left ventricular failure. After 24 hours, pulmonary edema had resolved, and the PIP and PaO2 had both returned to baseline. These data indicate that reperfusion of the ischemic lower torso leads to the synthesis of TxA2, an event temporally related to pulmonary hypertension and transient leukopenia with subsequent pulmonary microvascular injury manifest by interstitial edema.

Limb ischemia-induced increase in permeability is mediated by leukocytes and leukotrienes

This study tests the role of white blood cells (WBC) and leukotrienes in mediating the increased microvascular permeability following ischemia and reperfusion. Anesthetized dogs (n = 23) underwent 2 hours of hind limb ischemia induced by tourniquet inflation to 300 mmHg. In untreated animals (n = 7), tourniquet release led after 5 minutes to a rise in plasma thromboxane (Tx) B2 levels from 360 to 1702 pg/ml (p less than 0.05); after 2 hours, lymph TxB2 concentration had risen from 412 to 1598 pg/ml (p less than 0.05). There were decreases in circulating WBC from 11,766 to 6550/mm3 and platelets from 230 to 155 x 10(3)/mm3. During reperfusion, popliteal lymph flow (QL) increased from 0.07 to 0.24 ml/hour (p less than 0.05), while the lymph/plasma (L/P) protein ratio was unchanged from 0.39, changes consistent with increased microvascular permeability. WBC depletion (n = 7) to 302/mm3 by hydroxyurea or nitrogen mustard attentuated (p less than 0.05) the reperfusion induced rise in plasma TxB2 from 91 to 248 pg/ml and prevented the increase in lymph TxB2 concentration. Within 5 minutes of tourniquet release WBC counts further decreased to 191/mm3 (p less than 0.05) and platelets declined from 175 to 93 x 10(3)/mm3 (p less than 0.05). QL increased from 0.07 to 0.12 ml/hour (p less than 0.05), lower than untreated animals (p less than 0.05), and the L/P protein ratio declined from 0.49 to 0.37 (p less than 0.05), dilutional changes consistent with increased filtration pressure but not permeability to protein. Pretreatment with the lipoxygenase inhibitor diethylcarbamazine (DEC) (n = 8) prevented the reperfusion-induced increase in plasma and lymph TxB2 levels (p less than 0.05) and the fall in WBC counts (p less than 0.05), while platelet counts declined from 381 to 210 x 10(3)/mm3 (p less than 0.05). QL rose from 0.09 to 0.23 ml/hour (p less than 0.05) during reperfusion, and the L/P protein ratio of 0.3 remained unchanged, a value lower than in untreated dogs (p less than 0.05). In two animals of each group, vascular recruitment was induced by tourniquet inflation to 50 mmHg. This led to a high QL of 0.25 ml/hour and a low L/P ratio of 0.18. In untreated animals during reperfusion, QL further increased to 1.3 ml/hour, and L/P ratio rose to 0.44, documenting increased vascular permeability. In contrast, reperfusion in leukopenic or diethylcarbamazine (DEC)-treated dogs with vascular recruitment, was not associated with increases in QL or the L/P protein ratio.(ABSTRACT TRUNCATED AT 400 WORDS)

Thromboxane A2 mediates increased pulmonary microvascular permeability following limb ischemia.

Lower torso ischemia and reperfusion lead to respiratory dysfunction characterized by pulmonary hypertension and increased lung microvascular permeability. This is associated with lung leukose- questration and thromboxane (TX) generation. This study tests the role of elevated TX levels following muscle ischemia in mediating remote lung injury. Anesthetized sheep prepared with chronic lung lymph fistulae underwent 2 hours of bilateral hind limb tourniquet ischemia. In untreated controls (n=7), 1 minute after reperfusion there was a transient increase in plasma immunoreactive (i)-TXB2 levels from 211 to 735 pg/ml (p<0.05), and at 30 minutes, lung lymph i-TXB2 levels rose from 400 to 1,005 pg/ml (p<0.05). At 1 minute, the mean pulmonary arterial pressure (MPAP) increased from 13 to 38 mm Hg (p<0.05) and pulmonary microvascular pressure (Pmv) from 7 to 18 mm Hg (p<0.05). Lung lymph flow (QL) rose from 4.3 to 8.3 ml/30 min (p<0.05), the lymph/plasma (L/P) protein ratio was unchanged from 0.6, and the lymph protein clearance increased from 2.6 to 4.6 ml/30 min (p<0.05). Two hours after reperfusion, neutrophils were observed sequestered in lung capillaries and protemaceous exudates were found in alveoli in contrast to sham-operated animals (n=3). To maximize lung vascular surface area and acineve a pressure independent L/P protein ratio a left atrial balloon was inflated during one group of ischemia-reperfusion experiments (n=5). This resulted in a baseline rise in MPAP to 20 mm Hg (p<0.05); a 4.3-fold increase in QL (p<0.05), a decrease in the L/P ratio from 0.70 to 0.28 (p<0.05) and a protein reflection coefficient (ód) of 0.72. During reperfusion the L/P ratio rose to 0.49 (p<0.05) and the ód decreased to 0.51 (p<0.05), documenting an increase in lung microvascular permeability. In contrast to untreated ischemk controls, ininbition of TX synthetase with OKY 046 (n=6) reduced plasma i-TXB2 levels to 85 pg/ml (p<0.05) but also increased i-6-keto-PGF1α levels to 78 pg/ml relative to 15 pg/ml in untreated controls (p<0.05). OKY 046 prevented the increase in MPAP, Pmv, QL, and lymph protein clearance (p<0.05). Lung histology was normal in distinction to the leukosequestration in untreated ischemic controls. Pretreatment with OKY 046 combined with ibuprofen (n=5) prevented the increase in i-6-keto-PGF1α (p<0.05) but still led to a response unchanged from OKY 046 treatment alone. Pretreatment with the TX receptor antagonist SQ 29,548 (n=5) did not affect the ischemia induced increases in TXB, levels in plasma and lung lymph to 702 and 789 pg/ml, respectively, but prevented the increase in MPAP, Pmv, QL, lymph protein clearance, and lung leukosequestration (p<0.05 for all). These data indicate that the increased lung permeability following lower torso ischemia and reperfusion may be mediated by TX.

Role of neutrophil adherence receptors (CD 18) in lung permeability following lower torso ischemia.

Ischemia and reperfusion of the lower torso lead to leukotriene- and neutrophil (PMN)-dependent lung injury characterized by lung PMN sequestration, increased permeability, and noncardiogenic edema. It is thought that PMNs require adhesion to endothelium to alter barrier function. This study tests the role of CD 18, the PMN adherence receptor, in mediating lung permeability after lower torso ischemia and reperfusion. Anesthetized rabbits (n = 9) underwent 3 hours of bilateral hind limb ischemia. Ten minutes after the release of the tourniquets, plasma leukotriene B4 levels increased to 395 +/- 85 pg/ml, higher than 129 +/- 35 pg/ml in controls (n = 9, p less than 0.01). At this time there was a reduction in circulating white blood cells (x 10(3)), 3.56 +/- 0.49/mm3 relative to 6.07 +/- 0.61/mm3 in controls (p less than 0.01). PMNs were sequestered in the hind limbs, indicated by increased myeloperoxidase activity of 1.06 +/- 0.19 units/g compared with 0.56 +/- 0.09 units/g in controls (p less than 0.05). Four hours after tourniquet release, PMNs were sequestered in the lungs, 52 +/- 4 PMNs per 10 high-power fields, a value higher than 31.5 +/- 3 PMNs per 10 high-power fields in controls; bronchoalveolar lavage fluid protein content increased to 554 +/- 90 micrograms/ml relative to 277 +/- 46 micrograms/ml in controls; and there was lung edema, measured by increased wet weight-to-dry weight ratios of 5.19 +/- 0.10, higher than 4.29 +/- 0.21 in controls (all p less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)

Leukotrienes but not complement mediate limb ischemia-induced lung injury

Reperfusion after limb ischemia leads to sequestration of polymorphonuclear leukocytes (PMN) in the lungs and to leukocyte- (WBC) and thromboxane- (Tx) dependent respiratory dysfunction. This study examines the intermediary role of the chemoattractants leukotriene (LT)B4 and complement (C) fragments. Anesthetized sheep with chronic lung lymph fistulae underwent 2 hours of tourniquet ischemia of both hind limbs. In untreated controls (n = 7), 1 minute after tourniquet release, mean pulmonary artery pressure (MPAP) rose from 13 to 38 mmHg (p less than 0.05) and returned to baseline within 30 minutes. Pulmonary artery wedge pressure was unchanged from 3.6 mmHg. There were increases in plasma LTB4 levels from 2.46 to 9.34 ng/ml (p less than 0.01), plasma TxB2 levels from 211 to 735 pg/ml (p less than 0.05), and lung lymph TxB2 from 400 to 1005 pg/ml (p less than 0.05). C3 levels were 96% of baseline values. Thirty minutes after reperfusion, lung lymph flow (QL) increased from 4.3 to 8.3 ml/30 minutes (p less than 0.05), lymph/plasma protein ratio was unchanged from 0.6, and the lymph protein clearance increased from 2.6 to 4.6 ml/30 minutes (p less than 0.05), data consistent with increased microvascular permeability. WBC count fell within the first hour from 6853 to 3793/mm3 (p less than 0.01). Lung histology showed leukosequestration, 62 PMN/10 high-power fields (HPF) and proteinaceous exudates. In contrast to this untreated ischemic group, animals treated with the lypoxygenase inhibitor diethylcarbamazine (n = 5) demonstrated a blunted reperfusion-induced rise in MPAP to 17 mmHg (p less than 0.05). There were no increases in LTB4, TxB2, QL or lymph protein clearance (p less than 0.05). WBC count was unchanged and lung leukosequestration was reduced to 40 PMN/10 HPF (p less than 0.05). Decomplementation with cobra venom factor (n = 4) resulted in plasma C3 levels, 10% of baseline, but tourniquet release still led to pulmonary hypertension, elevated LTB4, TxB2 levels, and a decline in WBC count similar to that of untreated ischemic control animals. Histology showed 46 PMN/10 HPF sequestered in the lungs. Further, bilateral hind limb ischemia in either genetically sufficient (n = 10) or deficient (n = 10) C5 mice led to significant lung leukosequestration of 108 and 106 PMN/10 HPF, respectively, compared with 42 and 47 PMN/10 HPF in sham C5(+) and C5 (-) mice (n = 20) (p less than 0.01). These results suggest that the lung leukosequestration and increased microvascular permeability after lower torso ischemia are mediated by the chemotactic agent LTB4, but not by the complement system.

Vasodilating prostaglandins attenuate ischemic renal injury only if thromboxane is inhibited

Ischemia-induced renal injury is prevented by inhibition of thromboxane (Tx) synthesis. This protection was believed to be secondary to a high prostaglandin (PG)/TxA2 ratio. This study tests whether increasing the PG/Tx ratio by administration of vasodilating PGs protects the reperfused ischemic kidney. Anesthetized rats underwent right nephrectomy and 45 minutes of left renal pedicle clamping. Beginning 10 minutes before clamp release, animals were treated intravenously with the following: saline placebo (n = 10); the cyclooxygenase inhibitor ibuprofen (Ibu), 12.5 mg/Kg in a bolus (n = 8); a stable analogue of prostacyclin (PGI2), 500 ng/kg/minute for 2 hours (n = 9); PGE1, 400 ng/kg/minute for 2 hours (n = 8); the combination Ibu and PGI2 (n = 8) or PGE1 (n = 8). In saline treated ischemic controls, 5 minutes after reperfusion plasma, thromboxane (TxB2) and 6-keto-PGF1 levels were 2537 and 317 pg/ml, respectively--higher than the TxB2 and 6-keto-PGF1 levels of 750 and 80 pg/ml, respectively, in nephrectomized but nonischemic sham controls (n = 7) (p less than 0.05). In ischemic control animals at 24 hours, creatinine levels were 4.6 mg/dl, relative to 0.9 ml/dl in sham animals (p less than 0.05); the weight of the left (L) ischemic kidney relative to the right (R) normal kidney was 118%, compared with 99% in sham animals (p less than 0.05); and renal histology of ischemic control animals at 24 hours showed acute tubular necrosis (ATN) relative to normal findings in sham animals. Pretreatment with Ibu led to: TxB2 and 6-keto-PGF1 levels of 116 and 40 pg/ml, lower than those of sham animals (p less than 0.05); creatinine levels of 4.6 mg/dl, L/R renal weight of 119%; and ATN similar to that of ischemic controls. Treatment with a PGI2 analogue or PGE1 was not protective and led to increases in TxB2, 6-keto-PGF1, creatinine, L/R renal weight, and ATN similar to that of ischemic controls. The combination of Ibu and either PGI2 or PGE1 led to: reduced levels of TxB2 and 6-keto-PGF1 (p less than 0.05); attenuated increases in creatinine to 2.2 and 2.3 mg/dl, respectively (p less than 0.05); and limited ATN (p less than 0.05). These data indicate that the vasodilating PG protect the ischemic reperfused kidney only when Tx is inhibited.

Interleukin-2 induces early multisystem organ edema mediated by neutrophils

Interleukin-2 (IL-2), an agent known to activate neutrophils (PMN) with thromboxane (Tx)B2 release, produces pulmonary edema within 6 hours of intravenous infusion. This study tests the role of PMN in mediating the edema. Anesthetized rats received 106U recombinant human IL-2 (n = 15) or vehicle (n = 14) as a constant intravenous infusion during a period of 1 hour. At this time there was leukopenia 3.63 ± 0.43 (x103/mm3) relative to vehicle-infused control rats 6.12 ± 0.86 and a decline in PMN, 2.19 ± 0.14 relative to control value of 3.33 ± 0.05 (both p ≤ 0.05). After 6 hours edema, as measured by increase in the wet to dry weight (W/d) ratio, was present in the lungs (4.93 ± 0.20 relative to control 4.06 ± 0.10), heart (4.09 ± 0.11 versus 3.76 ± 0.08), liver (3.50 ± 0.10 versus 3.18 ± 0.10), and kidney (4.25 ± 0.07 versus 4.00 ± 0.07) (all p ≤ 0.05). There was increased lung permeability demonstrated by bronchoalveolarlavage fluid protein concentration of 1970 ± 210 Mg/mL relative to control 460 ± 90 μg/mL (p ≤ 0.05). Interleukin-2 resulted in lung PMN sequestration of 53 ± 7 PMN/10 high-power fields (HPF) relative to 23 ±2 PMN/10 HPF in controls (p ≤ 0.05) and increased plasma TxB2 levels to 1290 ± 245 pg/mL relative to control 481 ± 93 pg/mL (p ≤ 0.05). Pretreatment of other rats (n = 8) with selective anti-rat neutrophil antiserum 18 hours before the experiment led to a peripheral PMN count 10% of baseline and prevented edema in the lungs (W/d ratio 4.20 ± 0.16) and heart (3.67 ± 0.07) (both p ≤ 0.05) but not liver or kidney. Protein in lung lavage was reduced to 760 ± 220 ftg/mL (p ≤ 0.05). The protection afforded by leukopenia was associated with lack of PMN sequestration and prevention of the increase in plasma Tx levels (484 ± 120 pg/mL, p ≤ 0.05). These data indicate that the rapid induction of lung and heart edema with a 1-hour infusion of IL-2 in the rat is mediated, in large part, by activated PMNs.

Ischemia activates neutrophils but inhibits their local and remote diapedesis.

Hindlimb ischemia and reperfusion results in local limb and distant lung injury. This study tests whether the mechanism of injury is by ischemia mediated polymorphonuclear leukocyte (PMN) activation and diapedesis. Anesthetized rabbits were subjected to three hours of hindlimb ischemia (n = 8) or sham ischemia (n = 4). PMN derived solely from the reperfused ischemic limb and assayed flow cytometrically displayed an oxidative burst of 135 /- 8 fentamoles dichlorofluorescein (fmDCF)/cell compared to preischemc levels of 74 +/- 14 fmDCF/cell (p less than 0.05). Additional aliquots of isolated neutrophils were treated with phorbol myristate acetate (PMA) 10(-7) M. In contrast to a 162% increase in oxidative burst before ischemia, neutrophils at ten minutes of reperfusion had an enhanced response to PMA of 336% (p less than 0.05). Plasma collected from the ischemic hindlimb at ten minutes of reperfusion when introduced into an abraded skin chamber or intratracheally induced diapedesis in nonischemic animals. PMN accumulations in the skin chamber were 1636 +/- 258 PMN/mm3 after three hours (n = 8) compared to 63 +/- 18 PMN/mm3 induced by sham plasma (n = 4, p less than 0.05). Introduction of ischemic plasma intratracheally into a lobar bronchus (n = 4) induced PMN accumulations after three hours, measured by bronchoalveolar lavage fluid of 19 +/- 2 X 10(4) PMN/mm3 compared to 5 +/- 1 X 10(4) PMN/mm3 with sham plasma (n = 4, p less than 0.05). Diapedesis was completely prevented (0-3 PMN/mm3, p less than 0.05) by introducing ischemic plasma into skin chambers in animals whose hindlimbs had been made ischemic (n = 6) or into chambers located on skin regions that had been previously made ischemic (n = 6). Similarly after hindlimb ischemia, lavage of the lung with ischemic plasma yielded few PMN 0-3/mm3 (p less than 0.05). These data indicate that ischemia and reperfusion lead to generation of a circulating component in plasma that causes an oxidative burst in PMN and inhibits their diapedesis but promotes diapedesis when applied extravascularly to a naive animal.

The rapid induction by interleukin-2 of pulmonary microvascular permeability

The clinical use of interleukin-2 (IL-2) is limited by severe car-diopulmonary dysfunction. This study examines the mechanism of respiratory failure related to IL-2, using sheep with chronic lung lymph fistulae. Awake animals were infused with an intravenous (I.V.) bolus of IL-2 105 U/kg (n = 5) or its excipient (EXC) control (n = 3), every 8 hours for 4 to 5 days. Cardio-pulmonary function was monitored daily for at least one 8-hour period. Within 2 hours after each IL-2 administration, mean pulmonary arterial pressure (MPAP) rose. On Day 1, the mean rise was from 13 to 26 mmHg (p < 0.05), and on Day 5, to 29 mmHg (p < 0.05). MPAP returned to baseline levels after 2–3 hours. Pulmonary arterial wedge pressure was unchanged from 4 mmHg. There were transient falls in arterial oxygen tension, from 88 to 77 mmHg on Day 1 and to 73 mmHg (p < 0.05) on Day 5. Lung lymph flow (QL) rose from 2.4 to 6.8 ml/30 minutes (p < 0.05) on Day 1, and from 4.7 to 10.2 ml/30 minutes (p < 0.05) on Day 5, whereas the lymph/plasma protein ratio increased on Day 1 from 0.69 to 0.83 (p < 0.05) and from 0.63 to 0.71 (p < 0.05) on Day 5. This documents an increase in pulmonary microvascular permeability. Thromboxane (Tx)B2 levels increased transiently after each IL-2 injection in plasma from 195 to 340 pg/ml (p < 0.05) and in lung lymph from 222 to 772 pg/ml (p < 0.05) on Day 1, and to similar levels on Day 5. There was a progressive rise in cardiac output from 5.7 to 8.6 1/minute (p < 0.05) during the 5 days of infusion. Systemic blood pressure did not change. Temperature rose from 39.1 to 41.2 C (p < 0.05), and shaking chills were common. There was a progressive fall in leukocyte count, from 8.4 to 3.2 × 103/mm3 (p < 0.05) by Day 5, reflecting a 77% fall in lymphocytes. Lung lymph lymphocyte counts rose, and lymphocyte clearance increased. Lung histology revealed lymphocyte sequestration in capillary beds and peribronchial areas. EXC control infusions resulted in mild fever, increase in MPAP from 12 to 17 mmHg (p < 0.05), and QL from 2.6 to 5.1 ml/30 minutes, along with a fall in the L/P protein ratio from 0.77 to 0.71 (p < 0.05) changes that did not indicate increased permeability and were noted only on Day 1. The effect of IL-2 on endothelial cell (EC) barrier function was studied directly using bovine aortic EC grown to confluence on microcarrier beads. The EC barrier to the passage of trypan blue conjugated to albumin from the cultured medium into the microcarrier matrix was not altered by IL-2 (1–104 U/ml) or EXC. These data indicate that IL-2 leads to lung injury manifest by pulmonary hypertension and rapid increase in lung vascular permeability, effects likely mediated by lymphocytes and Tx. IL-2 alone has no direct effect on the vascular barrier.