“Imitators” of the ARDS*

Implications for Diagnosis and Treatment

Acute lung injury (ALI) and ARDS (ALI/ARDS)
are defined as follows: (1) an illness having an
acute onset, (2) an arterial oxygen
tension/inspired
oxygen fraction _ 200 mm Hg (or _ 300 mm Hg
for
ALI), (3) the presence of bilateral infiltrates on
frontal chest radiographs, and (4) a pulmonary
artery
occlusion pressure _ 18 mm Hg if measured, or
no
clinical evidence of left atrial hypertension
when not
measured.
There are a group of diffuse, noninfectious
parenchymal
lung diseases that often present in an acute
fashion and fulfill all of the clinical, physiologic,
and
radiographic criteria for ALI/ARDS. Some of
these
have distinct BAL characteristics and/or specific
histologic findings. It is important to distinguish
these from ALI/ARDS because most cases seem
to
respond to the early initiation of systemic
corticosteroid
therapy. In addition, studies evaluating novel
interventions for ALI/ARDS should not include
these patients because their outcomes are
dissimilar
from ALI/ARDS.
This opinion article summarizes what is known
about these diffuse noninfectious parenchymal
diseases
and focuses on the differences in their
histologies,
BAL findings, prognoses, and treatment
responsiveness.
We suggest that these conditions
should be considered, prompting further
diagnostic
testing, in every patient thought to have
ALI/ARDS
ascribed to “pneumonia,” and in those without
a
well-recognized predisposing factor.
Prevalence of Diffuse, Noninfectious
Parenchymal Lung Disease Mimicking
ALI/ARDS
In the absence of a prospective study of
consecutive
patients meeting the diagnostic criteria for ALI/
ARDS who undergo lung biopsy, it is not
possible to
know how frequently the acute diffuse,
noninfectious
parenchymal diseases meet the diagnostic
criteria
for ALI/ARDS. It is clear, however, that that
many patients with these conditions fit the
definition
of ALI/ARDS. The most common predisposing
causes of ALI/ARDS in most series are trauma,
aspiration, and sepsis. However, approximately
one third of the 861 patients enrolled in the
study3 of
low tidal volume ventilation for ALI/ARDS had
pneumonia identified as their precipitating
problem.
Specific diagnostic criteria for pneumonia were
not
defined. Rather, patients were assigned this
predisposing
factor on the basis of each investigator’s
clinical judgment at each site (R. Brower, MD;
personal communication; June 11, 2003). A
similar
prevalence of pneumonia is reported in other
series
of ALI/ARDS.
The diffuse, noninfectious parenchymal lung
diseases
often present with symptoms that are
consistent
with pneumonia. This may contribute to the
fact that despite extensive microbiologic and
serologic
testing, an infectious etiology can only be
found
in approximately 50% of patients presenting
with
what is considered to be a community-acquired
pneumonia. Accordingly, we suggest that
patients
thought to have ALI/ARDS on the basis of
pneumonia,
and those considered to have ALI/ARDS but
without a defined predisposing condition,
should
undergo BAL and, depending on the findings, a
lung
biopsy, to exclude one of the acute
noninfectious
parenchymal lung diseases.
The “Imitators”
Table 1 lists the histologic patterns, usual
etiologies,
and BAL cellular constituents of the acute
noninfectious parenchymal lung diseases that
often
meet ALI/ARDS criteria. These noninfectious
pneumonias
have protean manifestations, many of which
suggest an infectious process. These include
fever,
myalgias, nonproductive cough, and
progressive dyspnea.
9–10 The WBC counts, erythrocyte
sedimentation
rates, and C-reactive protein and serum lactic
dehydrogenase levels are often elevated.
Radiographic
studies often show diffuse alveolar infiltrates
reminiscent of ALI/ARDS. Although not always
the
case, the acute respiratory failure caused by
these
entities requires mechanical ventilation and
the use
of high levels of inspired oxygen and positive
end expiratory pressure as in ALI/ARDS.
Acute interstitial pneumonia (AIP) corresponds
to
the idiopathic clinico-pathologic entity
originally described
by Hamman and Rich. A similar presentation
and histology (ie, organizing diffuse alveolar
damage) representing one form of an acute
immunologic
pneumonia occurs with the collagen vascular
diseases (CVDs) or as the result of an
idiosyncratic
reaction to a pneumotoxic drug. AIP may
appear
in patients with preexisting idiopathic
pulmonary
fibrosis, and in this setting it has a particularly
dismal
prognosis. Idiopathic AIP (Hamman-Rich
syndrome)
is an acute, usually fulminant pneumonitis.
Patients have symptoms that have lasted from
a few
days to several weeks. The age range is wide,
the
mean being 50 years. Of the conditions listed
in
Table 1, AIP has the poorest outcome.
Persistence
and progression of fibrosis in surviving
patients, as
well as episodes of recurrence have been
documented.
17 In one series17 of 13 patients, only 1
patient had
full radiographic and physiologic recovery,
although
67% survived the initial hospitalization. One
report
indicates that seven of eight patients with AIP,
requiring mechanical ventilation and confirmed
by
surgical biopsies, survived following treatment
with
methylprednisolone (280 _ 143 mg/d). This was
compared to a 50% mortality in ARDS in the
same
study.18 AIP is the one acute noninfectious
parenchymal
disorder that is indistinguishable from ALI/
ARDS by BAL or surgical biopsy, and has been
referred to as ARDS without a precipitating
cause.
Clinical clues that suggest AIP as opposed to an
undefined pneumonia causing ARDS include
the
lack of accompanying multiorgan failure and
the
presence of cystic structures and traction
bronchiectasis
on the high-resolution CT scan, in addition to
the patchy ground-glass density and areas of
consolidation
that are expected in both AIP and ALI/ARDS.
Acute eosinophilic pneumonia (AEP) first
described
in 1989 usually occurs in healthy young
adults who present with a duration of illness
that is
usually less than a week and without a
discernable
precipitating event. Some cases, however,
have
occurred following environmental dust
exposure, the
initiation of cigarette smoking, or as a
noninfectious
complication of AIDS. A number of medications
can also result in acute eosinophilic lung
disease.
Peripheral eosinophilia, which is expected in
other
pulmonary eosinophilic syndromes is unusual
and, if
present, is not striking. Some patients have
high
fevers, pleuritic chest pain, and eosinophilic
pleural
effusions, the latter being a useful
distinguishing
feature in patients with diffuse infiltrates. All
patients
present with some degree of acute respiratory
failure and up to 50% require mechanical
ventilation.
Since AEP is a onetime event, and is very
sensitive to treatment, full recovery can be
expected.
There is a report of one patient in whom
AEP redeveloped after reinstitution of
tetracycline
therapy.
Bronchiolitis obliterans organizing pneumonia
(BOOP) is best recognized as a subacute
syndrome
process producing symptoms over several
months,
often following an antecedent upper
respiratory tract
infection. Less commonly, BOOP may be an
acute,
fulminant process resulting in respiratory
failure
requiring mechanical ventilation. The acute
form is either idiopathic or associated with a
CVD or
a drug toxicity. Imaging studies indicate patchy
nonspecific areas of consolidation. Although
the
existing literature only describes approximately
20
cases, the results of treatment for idiopathic
acute
BOOP seem to be inferior to that of subacute
BOOP. It is less likely to return with
corticosteroid
withdrawal as is the case in the subacute form
of
BOOP.
Diffuse alveolar hemorrhage (DAH) has many
causes and several possible underlying
histologies.
A confounding problem is that hemoptysis,
the expected symptom, is absent in up to 33%
of
cases. It is important to establish the cause of
DAH since virtually all cases, except those
associated
with overwhelming diffuse alveolar damage,
are potentially
reversible. Recurrences are to be expected
particularly if the underlying histology indicates
pulmonary
capillaritis (ie, a small vessel vasculitis of the
lung), the pathology that underlies some
systemic
vasculitidies; CVD; the antiphospholipid
antibody
syndrome; or anti-basement membrane
antibody
(ABMA) disease. Bland pulmonary hemorrhage,
defined as alveolar filling with RBCs, but
without septal inflammation, can result from
anticoagulant
overuse, thrombolytic therapy, and conditions
that result in thrombocytopenia.
Acute hypersensitivity pneumonitis (HP) is
characterized
by fever, cough, dyspnea, and an increased
WBC count appearing 4 to 6 h following the
inhalation
of an organic antigen. In most cases, it is a
recurrent, self-limited episode of pneumonitis
that is
often initially treated as a community-acquired
pneumonia until the nature of the disease is
discovered
and the antigen can be eliminated or avoided.
On occasion, acute HP can present with
fulminant
acute respiratory failure, which is clinically and
radiographically indistinguishable from
ALI/ARDS
or its other imitators. If unrecognized, cases of
fulminant acute HP can be fatal.

Evaluation of the Imitators

In patients who present with an acute,
noninfectious,
parenchymal lung disease without a recognized
ALI/ARDS risk factor, the possibility of an
illicit or prescribed drug, a systemic
immunologic
disease, or an environmental exposure being
the
cause must be considered along with infection.
Various
classes of drugs, particularly chemotherapeutic
agents, can manifest as any of the histologic
categories
in Table 1. An acute immunologic pneumonia
complicating a CVD can present with
organizing
diffuse alveolar damage, as in AIP, acute BOOP,
or
DAH secondary to either pulmonary capillaritis
or
bland hemorrhage. This can occur in a patient
with a previously diagnosed CVD, or the
pneumonitis may represent the initial
manifestation.
Although the lung may be the only apparent
organ involvement, a careful search for
systemic
involvement should be undertaken. It is critical
to
carefully seek evidence suggesting renal
involvement
in the form of a focal, segmental, necrotizing
glomerulonephritis
that is characterized by proteinura, hematuria,
and RBC casts. This rapidly progressive
glomerulonephritis is common to the CVD, the
systemic vasculitidies, and ABMA disease and
requires
prompt treatment if chronic renal insufficiency
is to be avoided.
Serologic testing directed at the
aforementioned
systemic diseases should be undertaken. In
addition,
serum creatinine phosphokinase and ferritin
should
be measured as increased levels may suggest
polymyositis
and adult-onset Still disease, respectively.
A reduced hematocrit in the face of diffuse
pulmonary infiltrates may be the first indication
of DAH.
Role of BAL
BAL should be performed early on and
preferably
after tracheal intubation. The samples should
be sent
for identification of selected infectious agents
and a
differential WBC count (Table 1). In addition to
excluding the possibility of the occasional
diffuse
fulminant viral or atypical pneumonia due to
Mycoplasma
pneumoniae, Legionella pneumophilia or
Chlamydia pneumoniae in the
immunocompetent
host, the results of the differential WBC count
may
obviate the need for further intervention. BAL
eosinophilia
in this setting establishes the diagnosis of
AEP again (peripheral eosinophilia is unusual).
Lymphocytosis without a marked elevation of
neutrophils
strongly suggests acute HP, although acute
BOOP may also demonstrate BAL
lymphocytosis
exceeding 30%. The finding of bloody aliquots
on sequential BAL samples indicates DAH. This
finding will require further evaluation as
outlined in
the previous section. Similar to ALI/ARDS and
the
infectious pneumonias, the BAL in AIP and
acute
BOOP can have a predominance of neutrophils
and
minor increases in the percentages of
lymphocytes
and eosinophils.

Timing and Indication for Surgical Biopsy

If the BAL results are nonspecific and the
clinical
condition of the patient permits, we
recommend
consideration for surgical biopsy. Most can be
performed
via the video-assisted transthorascopic
approach.
In patients with abnormal coagulation study
findings or thrombocytopenia, an open
thoracotomy
may be preferable. The open approach may
also be
appropriate if the degree of respiratory failure
is
severe. For each patient, the benefit vs risk of
surgical biopsy must be taken into
consideration.
Lung biopsy is recommended in patients with
DAH unless the etiology is obvious (eg, a
previously
diagnosed CVD, clinical features of a vasculitis
such
as Wegeners granulomatosis, microscopic
polyangiitis,
or Henoch-Schoenlein purpura that presents
with systemic manifestations in addition to
DAH and
pulmonary capillaritis). Moreover, pulmonary
capillaritis causing DAH may represent an
isolated
small-vessel vasculitis of the lung without
serologic
or systemic manifestations pointing to a CVD
or
vasculitis. DAH complicating a CVD may have
several underlying histologies, each with
different
treatment recommendations and prognoses.
Regardless
of the underlying histologies, it is important to
obtain immunofluorescent staining of lung
tissue in
this acute setting, since the presence of
immune
complexes is suggestive of a CVD, and the
presence
of linear basement membrane staining
indicates
ABMA disease.
Treatment of the Imitators
Data regarding the effectiveness of treatment
of
the acute, noninfectious parenchymal lung
diseases
come from individual reports and small case
series.
The literature also suffers from considerable
variability
in the timing, dosing, and duration of
treatment.
In general, corticosteroids are recommended.
In
addition, depending on the specific disorder,
immunosuppressive
drugs are added. Plasmapheresis is
recommended for patients with ABMA disease.
There are _ 90 cases of AIP recorded in the
literature, but the response to corticosteroid
treatment
can only be assessed in 26. For these, the
average mortality exceeded 70%. In a small
treated
group, 70% survived the acute event. In a
recent
series of 13 patients, the mortality was 50%;
even
allowing for the fact that 2 of these patients
died
after a recurrence of their AIP, and 2 others
acquired
progressive fibrotic lung disease, the mortality
in AIP
exceeds that currently attributed to ARDS.64
Information
regarding timing of therapy was unavailable.
In general, 250 mg of methylprednisolone was
administered as a single daily dose or repeated
for a
total of 1g/d IV for 3 days followed by a gradual
taper. Given the high mortality and possibly an
improved outcome, corticosteroid therapy, as
outlined
above, should be initiated in a suspected case
prior to tissue conformation. No information is
available
concerning additional therapies for the
treatment
of AIP.
Corticosteroids are very effective for AEP. The
literature reports _ 50 treated patients.
Complete
resolution occurs within 7 days, and without
residual
physiologic impairment. AEP in the vast
majority of
cases does not recur.
In acute BOOP, because of the paucity of
reported
cases, the results of treatment are less clear.
The
mortality may be as high as 70%, particularly if
it
accompanies a CVD. In one report, three of
five patients with acute idiopathic BOOP
survived
after receiving corticosteroids within the first
week
after biopsy. Corticosteroid treatment was
delayed
up to 2 weeks in the nonsurvivors.35 Other
studies support improved survival with early
diagnosis
and initiation of corticosteroid therapy.
DAH secondary to a systemic vasculitis, or a
CVD,
should be immediately treated with
corticosteroids
and a second immunosuppressive agent such
as
cyclophosphamide. Active glomerulonephritis is
an
additional mandate for immediate therapy.
Although
recurrences are possible as the medications
are
tapered, survival after the initial event
approximates
70% in patients with CVD, and 80% in those
with a
systemic vasculitis or isolated pauci-immune
pulmonary
capillaritis.
It is difficult to draw any conclusion regarding
the
efficacy of treatment for patients with acute
fulminant
HP, as only a few fatal cases appear in the
literature. Interestingly, all of these were
attributed
to avian antigens.

Summary
When confronted with a case of acute
respiratory
failure severe enough to meet the diagnostic
criteria
of ALI/ARDS but without a predisposing cause,
it is
important to consider the alternate diagnoses
reviewed
above, and to attempt to establish a diagnosis
as expediently as possible utilizing BAL with
differential
counts and, if necessary, the consideration for
surgical biopsy. The most common error is that
the
episode of respiratory failure is incorrectly
attributed
to an infectious pneumonia and appropriate
therapy
is either not administered or delayed.
Because the literature suggests that a delay of
specific therapy for some of these conditions
worsens
outcome, it seems prudent to administer
systemic
corticosteroids between 250 mg and 100 mg of
IV
methylprednisolone for a minimum of 3 days
prior to
the return of BAL, the microbiologic studies,
and/or
the surgical biopsy. If the microbiologic studies
reveal a potential causative agent,
corticosteroids
would obviously be discontinued.
Some would argue that any diffuse
pneumonitis
that fulfills the ALI/ARDS criteria is ALI/ARDS
and
should be treated accordingly.

Table 1—Acute Noninfectious Diffuse

Parenchymal
Lung Diseases and Their Underlying
Histology,
Etiology, and BAL Cellular Content
AIP
Organizing diffuse alveolar damage
Idiopathic (Hamman-Rich syndrome), CVD,
cytotoxic drugs,
infections
Neutrophilia (10%)
AEP
Eosinophilic infiltration and diffuse alveolar
damage
Idiopathic, drugs
Eosinophilia (25%)
Acute BOOP
Organizing pneumonia,
Idiopathic, CVD, drugs, radiation, infections
Neutrophilia, and sometimes lymphocytosis
(25%),
eosinophilia (25%)
DAH
Pulmonary capillaritis, bland hemorrhage,
diffuse alveolar
damage
Vasculitides, CVD, ABMA disease,
coagulopathies, diffuse
infections (partial list)
RBCs, hemosiderin-laden macrophages
Acute HP
Granulomatous and cellular pneumonitis with
diffuse alveolar damage
Environmental and workplace antigens
Lymphocytosis (25%) and sometimes
neutrophilia (10%)