Learning Center
Plans & pricing Sign in
Sign Out

High-Altitude Pulmonary Edema


									High-Altitude Pulmonary Edema
Characteristics of Lung Lavage Fluid
Robert B. Schoene, MD; Peter H. Hackett, MD; William R. Henderson, MD; E. Helene Sage, PhD;
Marguerite Chow; Robert C. Roach, MS; William J. Mills, Jr, MD; Thomas R. Martin, MD

To evaluate the cellular and biochemical composition of bronchoalveolar fluid in                         The low incidence of HAPE, the
high-altitude pulmonary edema (HAPE), we performed bronchoalveolar lavage                              remote locations where it occurs, and
in three climbers with HAPE in a research facility at 4400 m on Mount McKinley.                        the lack of a reproducible animal model
Three healthy climbers were used as controls. The HAPE fluids contained                                have made it difficult to study its
marked increases in high-molecular-weight proteins, erythrocytes, and leuko-                           pathophysiologic mechanisms. The
cytes, most of which were alveolar macrophages. The HAPE fluids also
                                                                                                       analysis of bronchoalveolar lavage
                                                                                                       fluid (BALF) has been used to charac¬
contained detectable amounts of leukotriene B4 and other lipoxygenase                                  terize a number of processes in the
products of arachidonic acid metabolism, complement fragments (C5a),                                   lung.11 We thought that this technique
inhibitors of neutrophil chemotaxis, and acid proteases but not hydroxyproline,                        could be safely used at a high-altitude
a constituent of collagen. The data from this study indicate that HAPE involves a                      research facility at 4400 m on Mount
transient "large pore" leak in the pulmonary circulation. Despite the presence                         McKinley, where cases of HAPE had
of two potent mediators of inflammation, leukotriene B4 and C5a, HAPE is not                           been seen during previous spring
characterized by the intense neutrophil accumulation that is typical of other                          climbing seasons. Our goals were to
forms of acute lung injury.                                                                            define the cellular and biochemical
                                                                 (JAMA 1986;256:63-69)                 composition of BALF from patients
                                                                                                       with HAPE and control subjects lav-
                                                                                                       aged at altitude and to compare the
HIGH-ALTITUDE                pulmonary edema               ment. For   example, HAPE has been          findings with other forms of pulmo¬
                                                           associated with the congenital absence      nary edema.
(HAPE) is              form of altitude
                 a severe
illness that occurs in 1% to 2% of                         of one pulmonary artery,5 and HAPE-         SUBJECTS AND METHODS
individuals ascending higher than 3000                     susceptible patients have elevated pul¬     Laboratory Location
m.1,2 High-altitude pulmonary edema                        monary artery pressures with normal            We performed the studies in the
was originally described as pneumonia                      left ventricular function.6,7 Other stud¬
or congestive heart failure,3,4 but was                    ies have linked a blunted hypoxic venti-    University of Alaska Denali Medical
                                                                                                       Research facility on a glacial plateau at
finally recognized as a form of noncar-                                                                4400 m adjacent to the West Buttress
diogenic pulmonary edema in I960.1                                See also pp 36 and 81.               of Mount McKinley, a route traveled by
While the etiology of HAPE remains
                                                                                                       hundreds of climbers each year. Two
unclear, it has been suggested that                        latory response and relative hypoventi-     weatherports supplied with power
high pressures in the pulmonary vascu-                     lation with HAPE, and these could           from solar panels and a gasoline gener¬
lature may play a role in its develop-
                                                           contribute to the marked alveolar           ator functioned both as a research
                                                           hypoxia and pulmonary hypertension          laboratory and a medical facility for
   From the Departments of Medicine (Drs Schoene,
                                                           observed in the syndrome.8 Although         climbers with altitude illnesses. This
Hackett, Henderson, and Martin) and Biological Structure   increased pulmonary vascular pres¬          facility was in radio contact with Tal-
(Dr Sage and Ms Chow), School of Medicine, University      sures could lead to hydrostatic edema,      keetna and Anchorage, Alaska, from
of Washington, Seattle: the Medical Service, Seattle       the demonstration of fibrin thrombi,
Veterans Administration Medical Center (Dr Martin); and
                                                                                                       which air evacuation by helicopter or
the Department of High Latitude Studies, University of     hyaline membranes, and neutrophil           fixed-wing aircraft could be arranged.
Alaska, Anchorage (Drs Hackett and Mills and Mr            infiltration in postmortem lung sam¬        The investigators gained access to the
  Reprint requests to Respiratory Disease Division,        ples of HAPE victims9,10 suggests that      laboratory over a three- to five-day
ZA-62, Harborview Medical Center, 325 Ninth Ave,           a   high-permeability, inflammatory         period of time by trekking to the
Seattle, WA 98104 (Dr Schoene).                            process might also be important.            laboratory from an airstrip on the

                                                Downloaded from by guest on May 7, 2011
Kahiltna glacier at 2200     m   outside   cell counts, and a 1.0-mL sample was        ("neutral"), and a third aliquot was
Denali National Park.                      frozen for total cell counts. The           incubated at room temperature for two
                                           remainder was filtered through 0.22-        hours with no treatment ("control").
Subjects                                   /¿m filters and the filtrate was frozen     Acidic and neutral incubations were
   The six male subjects included three    in an underground igloo (temperature,       terminated by the addition of 90 ¿tL of
climbers with HAPE (age, 32.7 ±6.7         approximately -10°C). The samples           Laemmli buffer containing 50mM di¬
years) and three healthy volunteers        later were coded, packed in dry ice, and    thiothreitol, and the samples were ana¬
 (age, 25.3 ±0.9 years). The patients      flown to Seattle, where they were           lyzed by SDS-PAGE.15
with HAPE were identified when they        stored (at -70°C) until analyzed. All
presented at the research station with     analyses were completed without             Analysis of Arachidonic Acid
marked shortness of breath, tachycar¬      knowledge of subject identity. Total        Lipoxygenase Products
dia, cyanosis, arterial oxygen desatura-   cell counts were performed in a hema-          The lavage samples were adjusted to
tion (mean arterial oxygen saturation      cytometer on the aliquots of freshly        80% alcohol (volume per volume) and
[SaoJ, 56.7% ±10.7%), and crackles         frozen, unfiltered lavage fluid. Filtered   centrifuged at 3000 g for ten minutes at
audible in the right side of the thorax    lavage samples were thawed, and the         4°C; the pellets were discarded. The
that spread to the left side during        first two of three aliquots were com¬       clear supernatants were adjusted to
several hours of observation. All of the   bined from each subject and concen¬         20% alcohol by the addition of water
HAPE victims had ascended rapidly,         trated tenfold by positive-pressure fil¬    and underwent solid-phase extraction
taken sedatives at night to sleep (which   tration under nitrogen at 4°C, using a      using octadecylsilene columns. The
may result in further hypoxemia), and      filtration membrane with a 1000-            eluants were evaporated to dryness
awakened in the morning with short¬        molecular-weight limit. The logistics of    under nitrogen, resuspended in methyl
ness of breath. None had clinical signs    the experiment dictated that the sam¬       alcohol/water/acetic acid (75:25:0.01,
of heart failure. The subjects were        ples be frozen. It is possible that the     volume per volume per volume) and
judged to be clinically capable of         freezing and thawing may have re¬           underwent reverse-phase high-per¬
                                           sulted in lysis of some of the cells and    formance     liquid  chromatography
undergoing bronchoscopy, and the pro¬      an underestimation of the total cell
cedure was fully explained to them and                                                 (HPLC)   on a   4.6x250-mm octadecylsi¬
the leaders of their climbing teams. All   counts; however, the HAPE and the           lene column (5-/im particle size) with
                                           control samples were treated identical-     the same solvent at a flow rate of 1
gave informed consent as approved by
the Human Subjects Experimentation         ly.                                         mL/min. Peaks at 280 and 235 nm were
Committees of the Universities of                                                      rechromatographed by reverse-phase
                                           Analysis of Lavage Proteins                 and straight-phase HPLC, as previous¬
Washington and Alaska. Spirometry
was performed during and after recov¬        Total protein was measured by sul-        ly described.17 The peaks were identi¬
ery from HAPE. Because of the severi¬      fasalycilic acid precipitation.12 Albu¬     fied by their ultraviolet spectra and
ty of their illness, two of the subjects   min, IgG, IgA, IgM, aj-antitrypsin,         HPLC co-elution with authentic leuko¬
were treated overnight with oxygen
                                                                                       triene (LT) and mono-hydroxy-eicosa-
                                           ceruloplasmin, and C3 activator    were
before lavage; the third was studied       measured in the concentrated and            tetraenoic acid (HETE) standards. The
about four hours after symptoms            unconcentrated fluids by radial im-         chemotactic activity of peaks that re¬
began. The three control subjects had      munodiffusion.13 The presence of            chromatographed with LTB4 was mea¬
been living at altitudes of 4400 m or      the complement fragments C5a and            sured by the leading front assay, using
above for more than ten days, were         C5adesarg was measured by radioimmu-        human neutrophils as previously de¬
asymptomatic, and had Sao2 values in       noassay.14                                  scribed.18 Unextracted BALF fluids
the range of those of other healthy           Equal aliquots (30 ng) of lavage         also were examined for slow-reacting
climbers (84.1% ±3.1%).                    samples were dialyzed against 0.1N          substance (SRS) activity, using         the
                                           acetic acid at 4°C, lyophilized, and        guinea pig ileal assay.18
Lavage Procedure                           analyzed by sodium dodecylsulfate-
   During the lavage procedure, all sub¬   polyacrylamide gel electrophoresis          Measurement of
jects were monitored with an ear oxim-     (SDS-PAGE), using dithiothreitol as a       Chemotactic Activity
eter and electrocardiogram, and giv¬       reducing agent.15 Human serum albu¬           The chemotactic activity of each la¬
en low-flow oxygen to maintain SaOs at     min (HSA) and low-molecular-weight          vage fluid for normal human neutro¬
or above 90%. A flexible fiberoptic        protein standards were used as refer¬       phils and monocytes was measured by
bronchoscope was passed transorally        ences. To identify albumin breakdown        the modified Boyden technique in
and wedged in a subsegment of the          products in the lavage fluids, equal        microchemotaxis chambers, using ei¬
right middle lobe. Three 40-mL ali-        aliquots of normal and HAPE lavage          ther nitrocellulose membranes with
quots of sterile, pyrogen-free 0.89%       samples were analyzed by SDS-PAGE,          3.0-/tm pores (for neutrophils)   or   poly¬
sodium chloride were instilled and         and the proteins were transferred elec-     carbonate membranes with 5.0-/¿m
recovered with gentle suction. Approxi¬    trophoretically to nitrocellulose pa¬       pores (for monocytes).19 Human neu¬
mately 70% of the total instilled fluid    per,16which was then incubated with         trophils and monocytes were recovered
was recovered from each subject. All       antisera to HSA, followed by iodine         from peripheral blood by Ficoll-Hy-
subjects tolerated the procedure well      125-labeled staphylococcal protein A.       paque   density gradient centrifugation
and were observed over that night by         To investigate the presence of acidic     and dextran sedimentation, washed
any one of us (R.B.S., P.H.H., or          protease activity, a 10-uL aliquot of       twice, and resuspended in phosphate-
R.C.R.).                                   frozen, untreated HAPE BALF (50.7           buffered saline (PBS) (pH 7.2) at
                                           Mg)  was acidified by the addition of       lXlOVmL.20 Aliquots of concentrated
Lavage Fluid Preparation                   0.1N hydrochloric acid and left at room     or unconcentrated lavage fluid, alone
   The lavage samples were processed       temperature for two hours ("acidic"). A     or diluted with PBS, were added to the
immediately in the following manner.       second aliquot was acidified and imme¬      lower compartments (25-/¿L volume)
One drop of fluid was air-dried on a       diately diluted with 90 pL of Laemmli       and 50 ¿¿L of cells was added to the top
glass slide for analysis of differential   buffer containing 50mM dithiothreitol       compartments. For comparison, zymo-

                            Downloaded from by guest on May 7, 2011
san-activated human     serum was    used     Characterization of Bronchoalveolar Lavage Fluid in High-Altitude Pulmonary Edema (Mean ±
as a standard positive stimulus in each
assay, and PBS was used to test ran¬
dom migration. The chambers were                                                                                                       High-Altitude
                                                                                                          Control (n=3)           Pulmonary Edema (n=3)
incubated for two hours at 37°C in
humidified 5% carbon dioxide/air, and         Cells X104/mL(X107/L)
the filters were removed and stained            Red blood cells                                               1.1 ± 0.5               261.4 ± 243.4
with a modified Wright's stain. The             White blood cells                                             4.3 ± 1.4                 36.7 ± 16.5
results are expressed as the number of
cells that migrated completely through
                                                   Alveolar    macrophages                                    4.0 ±1.3                  28.4 ±12.7
                                                                                                         (92.7% ± 3.3%)*             (78.7% ± 3.5%)t
the filters in a 1.0-mm2 eyepiece grid in
ten consecutive microscopic fields                 Neutrophils                                                0.1 ±0.1                   2.6 ±1.1
                                                                                                          (3.0% ± 1.0%)               (8.0% ±2.7%)
(X54Ó). All of the chemotaxis assays
were performed and analyzed before                 Lymphocytes                                                0.3 ±0.1                   5.7 ±2.9
breaking the code on the specimens.                                                                       (4.3% ±2.3%)               (13.3% ±6.1%)
                                              Proteins, mg/dL (g/L)                                         11.7±2.3                  673.0±234.6t
Statistical Analysis                            Total   protein                                            (0.12±0.02)                 (6.73±2.35)
   The results for the control subjects         Albumin                                                       6.4 ±0.9                390.0 ±141.8t
and the patients with HAPE were                                                                            (0.06±0.01)                 (3.90±1.42)
compared by Student's t test, and sig¬          IgG                                                          0.5 ±0.1                   85.7±33.7t
nificance was determined by a P value                                                                      (0.01 ± 0.00)               (0.86 ± 0.34)
of less than .05.
                                                IgA                                                           0.3±0.1                   14.9±5.7f
RESULTS                                                                                                  (0.003 ±0.00)                 (0.15 ±0.06)
  All of the patients with HAPE and             IgM                                                            ND*                       5.3 ±0.9
the control subjects tolerated the la¬                                                                                                 (0.05 ±0.01)
vage procedure without side effects.            «,-Antitrypsin                                                 ND                       32.9 ±7.3
After   overnight   treatment with oxy¬                                                                                                (0.33 ±0.07)
gen, one of the subjects with HAPE              Ceruloplasmin                                                  ND                        6.3 ±1.7
descended to sea level and two                                                                                                         (0.06 ±0.02)
descended to below 3000 m for two to
four days. These two returned within            C3 activator                                                   ND                        1.3 ±0.3
                                                                                                                                       (0.01 ±0.00)
seven days of their illness to climb to
the summit (6194 m), without relapse            C6fragments, ng/mL (C5a and C5ad8sarg)                        <1.8§                     27.2±7.4t
of HAPE.
                                                'Values in parentheses indicate percentage of each cell type in the differential cell count.
                                                tP<05    vs control.
Cellular Content                                ÍND indicates not detectable.
   Although comparable amounts of               §C5 fragments were detected in all three high-altitude pulmonary edema lavages, but only one of three control
lavage fluid were recovered from the
patients with HAPE and the control
subjects, the HAPE fluids contained
more erythrocytes (red blood cells           and         lymphocytes,            5.2% ±8.7%            HAPE fluid was acidified by dialysis
[RBCs]) and leukocytes (Table). The          (0.052 ±0.087). In comparison, the cell                   against 0.1N acetic acid, only very low
total leukocyte counts were increased        recoveries from the control subjects at                   levels of intact HSA appeared in the
in patients with HAPE, primarily             altitude are similar to these normal                      fluid. Instead, a major band of Mr
because of    an increased recovery of       values at sea level. The patients with                    50 000 and several lower-molecular-
 alveolar macrophages. Although the          HAPE had marked increases in total                        weight bands were observed (HAPE
total number of alveolar macrophages         cells, and slight increases in the per¬                   column, Fig 1, A). These proteins either
was increased in the HAPE lavages,           centages of neutrophils and lympho¬                       were    absent      or   present in very low
the mean percentage of macrophages           cytes, compared with reference normal                     amounts in   similarly acidified control
was significantly reduced compared           values.                                                  lavage fluids. Purification of this major
with that of the control subjects                                                                     polypeptide Mr 50 000 by molecular
because of the increase in the percent¬       Protein Content                                         sieve chromatography, followed by
ages of neutrophils (range, 4% to 13%           The concentration of all of the mea¬                  peptide mapping by cleavage with chy-
[0.04 to 0.13]) and lymphocytes (range,      sured proteins were increased in                         motrypsin, suggested that it was a
2% to 23% [0.02 to 0.23]). Because of
the variability in the data and the
                                             HAPE fluid, including IgM, aranti-                       degradation product of HSA. In addi¬
                                             trypsin, ceruloplasmin, and C3 activa¬                   tion, immunoblotting using electro-
small sample size, the changes in neu¬       tor (Table). These proteins were not                     phoretic transfer and antibodies to
trophil and lymphocyte percentages           detected in concentrated control la¬                     HSA revealed that several of the low¬
did not reach statistical significance. In   vage fluids. Fragments of C5a and/or                     er-molecular-weight bands in the
our laboratory at sea level, reference       C5adesarg, which are markers of comple¬                  HAPE lavage fluid reacted with anti-
normal values obtained from the la¬          ment activation, were present in all of                  HSA antibodies (Fig 1, B), consistent
vage of 21 normal, nonsmoking volun¬         the HAPE subjects. Very low levels of                    with degradation of albumin. Degrada¬
teers are as follows (means ±SD):            these products were detected in only                     tion could be inhibited by the addition
total white blood cells, 3.5±2.0X10V         one of the three controls.                               of pepstatin to the BALF before dialy¬
mL (3.5 X 2.0 X 107/L); alveolar macro¬         The major protein in HAPE and                         sis at acidic pH. Fibronectin and
phages, 93.0% ±8.5% (0.93 ±0.085);           control lavage was albumin (HSA) (Mr                     hydroxyproline were not detected in
neutrophils, 1.8% ±0.5% (0.018 ± 0.005);     67 000) (Fig 1, A). However, when                        HAPE      or   normal     lavage samples.

                                     Downloaded from by guest on May 7, 2011
                                                               Fig 1.—A, Analysis of normal and high-altitude
                                                               pulmonary edema (HAPE) lavage proteins by
                                                               sodium dodecylsulfate-polyacrylamide gel elec-
                                                               trophoresis and immunoblotting. Human serum
                                                               albumin (HSA) and low-molecular-weight pro¬
                                                               tein standards (LMW) we used as references.
                                                               Asterisk indicates Mr 50 000 band in HAPE
                                                               bronchoalveolar lavage fluid (BALF). B, HAPE
                                                               and control lavage samples incubated with
                                                               antisera to HSA. Both normal lavage protein
                                                               and HSA standard reacted with anti-HSA anti-
                                                               sera, as detected by autoradiography. Several
                                                               of lower-molecular-weight components in HAPE
                                                               lavage sample, including Mr 50 000 species,
                                                               also reacted with anti-HSA antisera. C, Acidifi¬
                                                               cation of HAPE BALF. The HAPE BALF initially
                                                               contained intact HSA, which is stable at neutral
                                                               pH ("control"). When incubated at acidic pH,
                                                               degradation to a major fragment of Mr 50 000
                                                               occurred ("acidic"). Dialysis of HAPE BALF at
                                                               acidic pH resulted in activation of protease(s)
                                                               with activity toward HSA. This enzymatic activity
                                                               was not present in normal BALF (A).

                                                                  Arachidonic Acid
                                                                  Lipoxygenase Products
                                                                    Lipoxygenase products of arachi¬
                                                                 donic acid metabolism were detected in
                                                                 each of the HAPE fluids by reverse-
                                                                 phase and straight-phase HPLC (Fig
                                                                 2). The major products recovered were
                                                                 LTB4, 5-(S),12-(A)-6-íraws-LTB<, 5-(S)-
                                                                 12-(S)-6-trcms-LTB4, and 11-hydroxyei-
                                                                 cosatetraenoic acid (HETE). The sulfi-
                                                                 dopeptide leukotrienes LTC4, LTD4, and
                                                                 LTE4 were not detected in HAPE fluids
                                                                 by HPLC, and SRS activity was not
                                                                 detected using the guinea pig ileum
                                                                 assay.18 No arachidonate lipoxygenase
                                                                 products were isolated from the control
                                                                 lavage fluids (data not shown).
                                                                 Chemotactic Activity
                                                                    Whereas normal lavage fluids con¬
             280   nm                          235   nm          tained chemotactic activity for neutro¬
0.01                                                             phils and monocytes, the HAPE lavage
                                                                 fluids attracted monocytes but not neu¬
                                                                 trophils (Fig 3). Concentration of the
                                                                 lavage fluids by positive-pressure fil¬
                                                                 tration augmented monocyte but re¬
                                                                 duced neutrophil chemotaxis, sug¬
                                                                 gesting that one or more chemotactic
                                                                 inhibitors were present in the HAPE
                                                                 fluids. To determine whether proteins
                                                                 in HAPE fluids might inhibit neutro¬
                                                                 phil chemotaxis, we added IgG and IgA

                                                              Fig 2.—Reverse-phase high-performance liquid
                                                              chromatography of arachidonic acid lipoxygen¬
                                                              ase products in high-altitude pulmonary edema
                                                              (HAPE). Representative chromatogram of one
       u.                                                     of HAPE bronchoalveolar lavage fluids is
                                                              shown. Absorbance of eluting peaks was mea¬
                                                              sured at 280 and 235 nm. These peaks were
                                                              identified: I indicates 5-(S),12-(fl)-6-frans-LTB<;
                        Elution Time, min                     II indicates 5-(S),12-(S)-6-/rans-LB<; III indi¬
                                                              cates LTB<; and IV indicates 11-hydroxyeicosa-
                                                              tetraenoic acid.

            Downloaded from by guest on May 7, 2011
             100   -i
                                                            H Control
                                                            I—I HAPE                                                        t   P<.01
                                                                                                                            *   P<.001
                                                            •    P<.05     o
                                                                 P<.025   %     30
     m                                                                          20                         JL,                  X
                                                                                       -Û. Control
                                                                                        PBS           0.06 0.6 3.0     0.02 0.2 0.5
                                                   Concentrated                              BALF          +                     +            +
                            Fluid                      Fluid                                              igG                   IgA       lgA=0.2

             350                                                               350-|

             300                                                               300


             200 H
                                                                          « 250

                                                                               200 H
                                                                                                                Ll i1, i
             150                                                               150

             100                                                               100

                                                                                50 H

                                                                                        PBS Control    0.06 0.6 3.0     0.02 0.2 0.5       lgG=0.6
                        Unconcentrated             Concentrated                              BALF
                            Fluid                        Fluid                                              IgA                 IgA        IgA =0.2

Fig 3.—Chemotactic activity in high-altitude pulmonary edema (HAPE) and control lavage fluids. Compared with control fluids, HAPE fluids did not attract
normal peripheral blood neutrophils (top left), but did attract monocytes (bottom left). Concentration of fluids augmented these effects. To test whether
protein components of HAPE fluid could alter chemotaxis, IgG and IgA were added in varying concentrations to one of control fluids and chemotactic
activity of fluid was assayed. IgG and IgA alone and in combination, in concentrations approximating those in HAPE fluids, suppressed neutrophil
migration (top right) and stimulated monocyte migration (bottom right) toward control bronchoalveolar lavage fluid (BALF). IgG and IgA concentrations are
expressed in milligrams per deciliter. PBS indicates phosphate-buffered saline.

to a control lavage fluid and retested              occurs  in HAPE. We found evidence of                 fluid suggests that the integrity of the
its chemotactic activity (Fig 3). The               an  increase in the number of recovera¬               endothelial-epithelial barrier has been
IgG and IgA suppressed neutrophil                   ble alveolar macrophages, RBCs, and                   lost and that a "permeability" leak has
chemotaxis and stimulated monocyte                  neutrophils, and activation of the                    occurred. An increase in the percentage
chemotaxis. The addition of IgG and                 complement pathway and the lipoxy-                    of neutrophils in lavage fluid is charac¬
IgA together at concentrations approx¬              genase pathway of arachidonic acid                    teristic of acute inflammatory proc¬
imating those found in HAPE fluids                  metabolism in the lavage fluids of                    esses that alter permeability and is
greatly inhibited the chemotactic activ¬            patients with HAPE. These findings                    typical of patients with the adult respi¬
ity for neutrophils.                                were associated with a leak of high                   ratory distress syndrome (ARDS).21"23
                                                    concentrations     of   plasma-derived                The finding of both high protein and
COMMENT                                             large-molecular-weight proteins that                  cell concentrations in patients with
   The research station on Mount                    normally are almost entirely excluded                 ARDS studied in our laboratory23
McKinley offered a unique opportunity               from the alveolar space.                              allows a comparison between them and
to study HAPE victims at high altitude                 Different forms of pulmonary edema                 our HAPE subjects. In these studies,
and enabled us to characterize HAPE                 have been characterized by the pres¬                  ARDS and HAPE BALF had similar
lavage fluid from acutely ill patients.             ence or absence of high concentrations                numbers of total white blood cells so
The purpose of this study was to define             of protein in airway or lavage fluid                  that the differential cell counts truly
the nature of the leak and the biochem¬             and/or an inflammatory response, usu¬                 reflect differences in cell populations,
ical and cellular characteristics associ¬           ally reflected by the accumulation of                 ie, HAPE BALF has a predominance of
ated with it. The results indicate that a           neutrophils in lavage fluid. The pres¬                alveolar macrophages, while ARDS
complex change in the cellular and                  ence of high concentrations of high-                  BALF contains mostly polymorphonu-
biochemical components of lung fluids               molecular-weight proteins in lavage                   clear neutrophils. The findings that

                                         Downloaded from by guest on May 7, 2011
HAPE lavage fluid contains markedly          ties.                                        of the 6-íraws-isomers of LTB4 recov¬
increased proteins but only small ele¬          The results of the chemotaxis assays      ered could have been derived from
vations in total neutrophils and the         are consistent with the lavage cell          LTC4.
percentage of neutrophils, suggests          profiles. The HAPE lavage fluids stim¬         The   source   of the   lipoxygenase prod¬
that the early events in the pathogene-     ulated monocyte migration, but inhib¬        ucts in the HAPE BALF is unknown.
sis of HAPE involve a change in             ited neutrophil migration. Concentra¬        Stimulated human alveolar macro¬
permeability, without major activation      tion of the HAPE fluids augmented            phages produce LTB4       and its two 6-
of inflammatory mechanisms that lead        these effects. This suggests that, at this
to neutrophil accumulation in the
                                                                                         frems-isomers,19,42 and these cells are
                                            time in the clinical course of HAPE,         greatly increased in number in the
lung.                                       one or more inhibitors of neutrophil         HAPE fluid. Neutrophils17 and airway
   Catheterization studies at altitude      chemotaxis are present in HAPE flu¬          epithelial cells43 also generate LTB4 and
have indicated that HAPE is a noncar-       ids. It is possible that chemotactic         could contribute to the production of
diogenic form of pulmonary edema            inhibitors known to be present in plas¬      this mediator in the HAPE BALF. A
associated with high pulmonary artery       ma exude into the alveolar space             novel finding was the recovery of 11-
pressures and normal pulmonary capil¬       because of the change in vascular per¬       HETE in the alveolar lavage fluid of
lary wedge pressures.24 This finding        meability.31 The finding that IgG and        the patients with HAPE. The 11-HETE
has led to the suggestion that HAPE is      IgA suppressed neutrophil, but not           has weak chemotactic activity for
a vascular leak caused by overperfusion    monocyte chemotaxis, raises the possi¬        human neutrophils.44 The 11-HETE is a
of the pulmonary vascular bed. Exag¬       bility that these and other serum-            major product of cultured rat aorta
gerated hydrostatic forces secondary to    derived proteins also might function as       smooth-muscle cells45; 11-HETE pro¬
a    combination of an accentuated         local inhibitors of chemotaxis. Aggre¬        duction by vascular smooth-muscle
hypoxic pulmonary vascular response        gates of IgG have been found to inhibit       cells is inhibited by aspirin and indo-
and the increased cardiac output at        neutrophil migration,32,33 as have IgA        methacin, suggesting that it is a cyclo-
altitude were thought to lead to a         myeloma proteins.34 However, the dif¬         oxygenase product in these cells.45 It is
low-protein, hydrostatic type of leak.25   ferential effects of IgG and IgA on           unknown whether the 11-HETE recov¬
Before this study, however, this hy¬       neutrophil and monocyte chemotaxis            ered in the HAPE BALF similarly
pothesis had not been tested, as numer¬    that we have observed when these              represents   a   cyclo-oxygenase product.
ous attempts to create HAPE in an          proteins were added to normal lavage             The protein electrophoresis data
animal model have been inconsistent or     fluid have not been reported previous¬        indicate that the HAPE fluids also
unsuccessful.25"30                         ly, to the best of our knowledge. Recent      contained high concentrations of inac¬
  The present data indicate that           data also suggest that products of            tive acidic proteases. A possible source
HAPE is    a high-protein, high-permea¬    alveolar macrophages can inhibit neu¬         of the proteolytic activity is the alveo¬
bility type of pulmonary edema, as the     trophil chemotaxis, suggesting that           lar macrophage, which secretes a varie¬
protein concentrations are in the range    alveolar macrophages themselves               ty of proteases, including a plasmino-
 of those found in ARDS.22 The presence    might be the source of inhibitor sub¬         gen activator specific for extracellular
 of RBCs in HAPE BALF raises the           stances.35                                    matrix components.46,47 The rapid re¬
 possibility that areas of microhemorr-       Despite the relative lack of neu¬          covery of HAPE victims suggests that
 hage might be occurring within the        trophil accumulation, two potential           HAPE does not involve major disrup¬
 lung. However, the ratio of proteins to    mediators of inflammation were de¬           tion of lung parenchyma, which is
 RBCs was more than 100 times greater       tected in the HAPE fluids: C5 frag¬          characteristic of patients with intense
 than that found in blood for all of the    ments and LTB4. The presence of C5           inflammatory processes in the lung.48
 proteins measured, suggesting that         fragments (C5a and/or C5adesarg) in all         The cause of the injury to the pulmo¬
 microhemorrhage alone does not ex¬         of the HAPE fluids suggests that some        nary vasculature in HAPE remains
 plain the data.                            degree of complement activation had          uncertain, but it may occur in part
   Although the concentrations and size     occurred. Additionally, lipoxygenase         because of a sequence of events ini¬
 ranges of the proteins in the HAPE         products of arachidonic acid were pres¬      tiated by hypoxia and inordinately
fluids are similar to those found in        ent in all of the HAPE fluids but none       high pulmonary vascular pressures. It
ARDS, the predominance of alveolar          of the controls (Fig 2), and LTB4 was        is possible that "shear forces" from the
macrophages rather than neutrophils         the predominant product. The sulfido-        high intravascular pressure may lead
in the BALF is strikingly different        peptide LTs, which have SRS activity,         to increased permeability of the vascu¬
from the lavage cellular profile in        were not detected either by bioassay or       lar endothelium. Further studies will
ARDS.21"23 Possible explanations for       by HPLC. Among other properties,              be necessary to determine what causes
the greater recovery of alveolar macro¬    LTB4 has chemotactic activity for leu¬        the leak in this unique form of revers¬
phages in HAPE are an alteration of        kocytes and may be an important               ible pulmonary edema and what role
their adherence properties by edema in     mediator of inflammation in the lung.19       alveolar macrophages, neutrophils, and
the lung, with greater washout of cells    Leukotriene C4 and LTD4, which consti¬        inflammatory mediators might play in
that are normally present, or an           tute SRS, cause bronchoconstriction           its pathogenesis. The facts that HAPE
increased recruitment of blood mono-       and increase systemic vascular perme¬         victims improve rapidly and that two
cytes into the lung by chemotactic         ability.3637 Leukotriene C4 and LTD4          of our subjects climbed to the summit
factors generated in the alveoli. Al¬      have been recovered previously in air¬        within ten days indicate that the per¬
though the total number of lavage          way fluids of neonates with respiratory       meability increase in HAPE subjects
neutrophils was increased, their per¬      distress38 and from adults with               resolves rapidly and further differenti¬
centage was only slightly greater than     ARDS.39 The failure to detect LTC4 and        ates HAPE from other forms of acute
the normal range for lavage neutro¬        LTD4 in the HAPE lavage fluids may            lung injury.
phils. This finding also could be con¬     reflect either their true absence or
sistent with an increased washout of       their degradation before assay. As              This investigation was supported in part by
neutrophils normally present, perhaps      LTC4 is degraded oxidatively to the           grants HL30542 and HL234163 from the Public
                                                                                         Health Service; Clinical Investigator Grant HL-
because of altered adherence proper-       6-iraws-isomers of LTB4,17,40,41 a portion    00906 from the National Heart, Lung, and Blood

                            Downloaded from by guest on May 7, 2011
Institute, Rockville, Md, and   an E. L. Trudeau          Immunochemical quantitation of antigens by sin-             31. Kreutzer DL, Claypool WD, Jones M, et al:
Scholar Award from the American Lung Associa¬             gle radial immunodiffusion. Int J Immunochemis-             Isolation by hydrophobic chromatography of the
tion to Dr Schoene; and Allergic Diseases Academ¬         try 1965;2:235-254.                                         chemotactic factor inactivators from human
ic Award AI00487 from the National Institute of           14. Hugli TE, Chemoweth DE: Biologically active             serum. Clin Immunol Immunopathol 1979;12:162\x=req-\

Allergy and Infectious Diseases to Dr Henderson.          peptides of complement: Techniques and signifi-             176.
Dr Sage is an established Investigator of the             cance  of C3a and C5a measurements. Lab Res                 32. Wilkinson PC: Effects of human IgG on
American Heart Association, with funding in part          Methods Biol Med 1980;4:443-460.                            locomotion of human neutrophils related to IgG
from its Washington Affiliate.                            15. Laemmli UK: Cleavage of structural proteins             binding of a hydrophobic probe. Immunology
   The LT standards were provided by J. Rokach,           during the assembly of the head of bacteriophage            1980;41:457-466.
Merck Frosst Laboratories, Pointe-Claire/Dorval,          T4. Nature 1970;227:680-685.                                33. Kay NE, Bumol TF, Douglas SD: Effects of
Quebec, Canada. The HETE standards were pro¬              16. Towbin H, Staehelin T, Gordon J: Electro-               phagocytosis and Fc receptor occupancy on
vided by W. C. Hubbard, National Institutes of            phoretic transfer of proteins from polyacrylamide           complement dependent neutrophil chemotaxis. J
Health, Bethesda, Md.                                     gels to nitrocellulose sheets: Procedure and some           Lab Clin Med 1978;91:850-856.
   We thank Karl Maret, Gertrude Chiang, Venus            applications. Natl Acad Sci USA 1979;79:4350\x=req-\        34. Van Epps DE, Williams R: Suppression of
Wong, Brian and Diane Okonek, Talkeetna Air               4354.                                                       leukocyte chemotaxis by human IgA myeloma
Taxi, K-2 Aviation (logistic support on Mount             17. Henderson WR, Klebanoff SJ: Leukotriene                 components. J Exp Med 1976;144:1227-1242.
McKinley), and Norma Jean Schwab for her                  production and inactivation by normal, chronic              35. Sibille Y, Naegel GP, Merrill WW, et al:
assistance with the manuscript.                           granulomatous disease and myeloperoxidase-defi-             Profile of chemotactic factors produced by acti-
   The study is dedicated to Zdenek Hermann, a            cient neutrophils. J Biol Chem 1983;258:13522\x=req-\       vated alveolar macrophages: A complex pattern
Czechoslovakian climber who died of HAPE in               13527.                                                      including an inhibitor substance, abstracted. Am
1983 in the Pamir Mountains of the Soviet Union.          18. Henderson WR, J\l=o"\rgA, Klebanoff SJ: Eosino-         Rev Respir Dis 1985;131:A388.
                                                          phil perioxidase mediated inactivation of leuko-            36. Lewis RA, Austen KF: The biologically active
                                                          trienes B4, C4, and D4. J Immunol 1982;128:2609\x=req-\     leukotrienes: Biosynthesis, metabolism, functions,
References                                                2613.                                                       and pharmacology. J Clin Invest 1984;73:889-897.
                                                          19. Martin TR, Altman LC, Albert RK, et al:                 37. Samuelson B: Leukotrienes: Mediators of
1. Houston CS: Acute pulmonary edema of          high     Leukotriene B4 production by the human alveolar             immediate hypersensitivity reactions and inflam-
altitude. N Engl J Med 1960;263:478-480.                  macrophage: A potential mechanism for amplify-              mation. Science 1983;220:568-575.
2. Hultgren H, Spickard W, Hillriegel K, et al:           ing inflammation in the lung. Am Rev Respir Dis             38. Stenmark KR, James SL, Voelkel NC, et al:
High altitude pulmonary edema. Medicine 1961;             1984;129:106-111.                                           Leukotrienes C4 and D4 in neonates with hypoxe-
40:289-313.                                               20. Boyum A: Isolation of mononuclear cells and             mia and pulmonary hypertension. N Engl J Med
3. Ravenhill TH: Some experience of mountain              granulocytes from human blood. Scand J Clin                 1983;309:77-80.
sickness in the Andes. J Trop Med Hyg 1913;               Invest 1968;21(suppl 97):77-89.                             39. Matthay MA, Eschenbacher WL, Goetzl EJ:
20:313-320.                                               21. Lee CT, Lewis RA, Lippman M, et al: Elasto-             Elevated concentrations of leukotriene D4 in pul-
4. Hurtado A: Aspectos fisiologicos y patologicos         lytic activity in pulmonary lavage fluid from               monary edema fluid of patients with the adult
de la vida en la altura. Lima, Peru, Editora Rimae,       patients with adult respiratory distress syndrome.          respiratory distress syndrome. J Clin Immunol
1937.                                                     N Engl J Med 1981;304:192-196.                              1984;4:479-483.
5. Hackett PH, Creagh CE, Grover RF, et al: High          22. McGuire W, Spragg RG, Cohen AB, et al:                  40. Goetzl EJ: The conversion of leukotriene C4 to
altitude pulmonary edema in persons without the           Studies on the pathogenesis of the adult respirato-         isomers of leukotriene B4 by human eosinophil
right pulmonary artery. N Engl J Med 1980;                ry distress syndrome. J Clin Invest 1982;69:543\x=req-\     peroxidase. Biochem Biophys Res Commun 1983;
302:1070-1073.                                            553.                                                        106:270-275.
6. Hultgren HN, Grover RF, Hartley LH: Abnor-             23. Maunder RJ, Martin TR, Moore R, et al: The              41. Lee CW, Lewis RA, Corey BJ, et al: Oxidative
mal circulatory responses to high altitude in             safety of bronchoalveolar lavage in patients with           inactivation of leukotriene C4 by stimulated poly-
subjects with a previous history of high altitude         the adult respiratory distress syndrome (ARDS),             morphonuclear leukocytes. Proc Natl Acad Sci
pulmonary edema. Circulation 1971;44:759-770.             abstracted. Am Rev Respir Dis 1985;131:A137.                USA 1982;79:4166-4170.
7. Viswanathan R, Subramanian S, Lodi STK, et             24. Hultgren HN, Lopez CE, Lundberg E, et al:               42. Fels AO, Pawlowski NA, Dramer EB, et al:
al: Further studies on pulmonary edema of high            Physiologic studies of pulmonary edema at high              Human alveolar macrophages produce leukotriene
altitude: Abnormal responses to hypoxia of men            altitude. Circulation 1964;29:393-428.                      B4. Proc Natl Acad Sci USA 1982;79:7866-7870.
who had developed pulmonary edema of high                 25. Hultgren HN: High altitude pulmonary ede-               43. Holtzman MJ, Aizawa H, Nadel JA, et al:
altitude. Respiration 1978;36:216-222.                    ma, in Staub NC (ed): Lung Water and Solute                 Selective generation of leukotriene B4 by tracheal
8. Hackett PH, Schoene RB, Roach RC, et al:               Exchange. New York, Marcel Dekker Inc, 1978,                epithelial cells from dogs. Biochem Biophys Res
Blunted chemosensitivity and hypoxic ventilatory          pp 437-469.                                                 Commun 1983;114:1071-1076.
depression in high altitude pulmonary edema,              26. Erdmann AJ III, Vaughan TR Jr, Brigham                  44. Goetzl EJ, Sun FF: Generation of unique
abstracted, in Sutton JR, Houston CS, Coates G            KL, et al: Effect of increased vascular pressure on         mono-hydroxy-eicosatetraenoic         acids   from
(eds): Hypoxia and Cold. New York, Prager Pub-            lung fluid balance in unanesthetized sheep. Circ            arachidonic acid by human neutrophils. J Exp
lishers, 1986.                                            Res 1975;37:271-284.                                        Med 1979;150:406-411.
9. Arias-Stella J, Kruger H: Pathology of high            27. Bland RD, Demling RH, Selinger SL, et al:.              45. Bailey JM, Bryant RW, Whiting J, et al:
altitude pulmonary edema. Arch Pathol 1963;               Effects of alveolar hypoxia on lung fluid and               Characterization of 11-HETE and 15-HETE
70:43-53.                                                 protein transport in unanesthetized sheep. Circ             together with prostacyclin as major products of
 10. Dickinson J, Heath D, Gosney J, et al: Altitude      Res 1977;40:269-274.                                        the cyclo-oxygenase pathway in cultured rat aorta
related deaths in seven trekkers in the Himalaya.         28. Ohkuda K, Nakahara K, Weidner WJ, et al:                smoothe muscle cells. J Lipid Res 1984;24:1419\x=req-\
 Thorax 1983;38:646-656.                                  Lung fluid exchange after uneven pulmonary                  1428.
11. Hunninghake GW, Gadek JE, Kawanami O, et              artery obstruction in sheep. Circ Res 1978;43:152\x=req-\   46. Goldstein E: Hydrolytic enzymes of alveolar
al: Inflammatory and immune processes in the              161.                                                        macrophages. Rev Infect Dis 1983;5:1078-1092.
human lung in health and disease: Evaluation by           29. Landolt CC, Matthay MA, Albertine RH, et al:            47. Chapman HA, Stone OL, Vairin Z: Degrada-
bronchoalveolar lavage. Am J Pathol 1979;97:149\x=req-\   Overperfusion, hypoxia, and increased pressure              tion of fibrin and elastin by intact human alveolar
206.                                                      cause only hydrostatic pulmonary edema in anes-             macrophages in vitro. J Clin Invest 1984;73:806\x=req-\
12. Muelemans O: Determination of total protein           thetized sheep. Circ Res 1983;52:335-341.                   815.
in spinal fluid with sulfasalicylic acid and trichlor-    30. Mitzner W, Sylvester JT: Hypoxic vasocon-               48. Raghu G, Striker L, Hudson LD, et al: Extra-
oacetic acid. Clin Chem Acta 1960;5:757-761.              striction and fluid filtration in pig lungs. J Appl         cellular matrix in normal and fibrotic human
13. Mancini G, Carbonara AO, Heremans JF:                 Physiol 1981;51:1065-1071.                                  lungs. Am Rev Respir Dis 1985;131:281-289.

                                             Downloaded from by guest on May 7, 2011

To top