The

n e w e ng l a n d j o u r na l

of

m e dic i n e

clinical problem-solving

A Crazy Cause of Dyspnea

Amy Leigh Miller, M.D., Ph.D., Scott Schissel, M.D., Ph.D., Bruce D. Levy, M.D.,
and Joseph Loscalzo, M.D., Ph.D.
In this Journal feature, information about a real patient is presented in stages (boldface type)
to an expert clinician, who responds to the information, sharing his or her reasoning with
the reader (regular type). The authors commentary follows.

From the Clinical Pathological Conference

Series, Department of Medicine, Brigham
and Womens Hospital and Harvard Medical School all in Boston. Address reprint
requests to Dr. Levy at Brigham and Womens Hospital, Harvard Institutes of Medicine Bldg., Ave. Pasteur (HIM855), Boston, MA 02115, or at blevy@partners.org.
N Engl J Med 2011;364:72-7.
Copyright 2011 Massachusetts Medical Society.

An Interactive
Medical Case
related to this
Clinical ProblemSolving article
is available at
NEJM.org

An 18-year-old black woman with a history of asthma presented with fever, ear pain,
and dull discomfort on the right side of her chest that was unchanged with movement
or inspiration. She had no other symptoms and had been well until the morning of
presentation. On examination, her temperature was 38.8C, and her right tympanic
membrane was inflamed. Although the lungs were clear on auscultation, radiography revealed air-space opacities in both lower lobes. Azithromycin was prescribed.
Her symptoms resolved within 24 hours, and repeat radiography performed 1 week
later showed that the opacities, although still present, had diminished.
In a healthy young adult presenting with acute fever and otitis media, it is most
likely that the observed pulmonary opacities are infectious in origin. Infection with
Streptococcal pneumoniae or Chlamydia pneumoniae is a common cause of both otitis
media and pneumonia. Incomplete radiographic resolution 1 week after presentation
is not surprising, since the radiographic features of pneumonia may take weeks to
fully resolve. In the absence of continued or recurrent symptoms, no additional
imaging would be necessary.
Several months later, the patient returned, reporting a nonproductive cough and dyspnea on exertion. Frequent coughing spells had forced her to discontinue participation in competitive soccer and dance. Initially, the dyspnea occurred only with coughing, but it slowly progressed to the point of limiting her functional capacity, despite
the use of inhaled albuterol. She had no fevers, chills, or night sweats, and her weight
was stable. The results of a physical examination were unremarkable. The patient was
referred for asthma education, through which potential home triggers (a cat and
wall-to-wall carpeting) were identified and poor inhaler technique was noted. It was
not possible to complete pulmonary-function testing because of the patients cough.
She was given a spacer and peak-flow meter, began using a fluticasonesalmeterol
inhaler (delivering 250 g of fluticasone and 50 g of salmeterol per puff), and was
given a prescription for loratadine in addition to the albuterol.
A logical question at this point is whether the patients current symptoms are related
to her initial presentation. Has the initial, presumably infectious process failed to
resolve? Is her current illness the result of her initial illness (e.g., postinfectious
bronchiolitis obliterans with organizing pneumonia)? A chest radiograph would
have been useful to see whether there are persistent or recurrent opacities. Alternatively, her cough and shortness of breath may reflect progression of her asthma;

72

n engl j med 364;1 nejm.org january 6, 2011

The New England Journal of Medicine

Downloaded from nejm.org on July 20, 2011. For personal use only. No other uses without permission.
Copyright 2011 Massachusetts Medical Society. All rights reserved.

clinical problem-solving

more information is needed on her asthma history. Although a course of relentless and progressive asthma symptoms is somewhat atypical at
her age, an asthma-plus syndrome, such as the
ChurgStrauss syndrome, chronic eosinophilic
pneumonia, parasitic infection, or allergic bronchopulmonary aspergillosis, is possible. An acquired or congenital immunologic defect should
be considered as a cause of possible recurrent
infection, including a humoral defect (e.g., hypo
gammaglobulinemia or hyper-IgM or hyper-IgE
syndrome), a phagocytic defect (e.g., chronic granulomatous disease), or a defect in cellular immunity (e.g., infection with the human immunodeficiency virus [HIV]). Alternatively, her asthma
could be a misdiagnosis of symptoms caused by
an underlying lung disease that has predisposed
her to infection (e.g., cystic fibrosis unlikely,
given her race or congenital bronchiectasis).
The patients asthma had previously occurred only
with exercise and had been well controlled when
she used albuterol before beginning activities; she
had never required glucocorticoids or hospitalization. Her medical history also included seasonal
allergic rhinitis, an allergy to bee stings, eczema,
and pneumothorax at birth. Her medications included albuterol, loratadine, and fluticasonesalmeterol. She had no known drug allergies. Her
family history was notable for an aunt with sarcoidosis and multiple family members with allergies and asthma. The patient was a high-school
student and lived with her mother. She had no history of alcohol, tobacco, or illicit drug use and had
not traveled recently.
Although sarcoidosis is less likely to be familial
in blacks than in whites, the patients family history raises the possibility of a diagnosis of sarcoidosis. Eczema and allergic rhinitis are commonly
associated with asthma and support a diagnosis
of asthma or an asthma-plus syndrome as the
cause of her symptoms. Exercise-induced dyspnea
and cough are common in patients with asthma
and thus are not helpful in narrowing the differential diagnosis. The absence of recurrent spontaneous pneumothorax makes it likely that the
cause of neonatal pneumothorax was isolated, perhaps owing to a birth-related trauma or positivepressure ventilation in the newborn period; it is
unlikely to be related to the current presentation.

Over the next 2 months, symptoms progressed to

the extent that the patient noted considerable dyspnea while walking and occasional dyspnea while
talking. She continued to have episodes of nonproductive cough, although they were infrequent.
She had occasional night sweats but no fevers or
chills. She reported having dry eyes and dry skin
but no pain or discomfort in her joints. Despite
having a normal appetite, the patient had lost
about 1.5 kg since her initial presentation. On repeat physical examination, she did not appear ill.
Her temperature was 36.7C, heart rate 86 beats
per minute, blood pressure 98/60 mm Hg, and
oxygen saturation 96% while she was breathing
ambient air. The conjunctivae were clear, and funduscopic examination was normal. The tympanic
membranes were normal, and the mucous membranes were moist. The neck was supple, without
thyromegaly, and there was no lymphadenopathy.
The jugular venous pressure was not elevated. The
lungs were clear on auscultation. Cardiac examination revealed a regular rate and rhythm, with no
murmurs, rubs, or gallops. The abdomen was soft
and nontender. The extremities were warm and
well perfused. The neurologic examination was
unremarkable.
The patients clinical status has worsened substantially, with progressive dyspnea, functional
limitation, and extrapulmonary symptoms, including symptoms consistent with a sicca-type syndrome, weight loss, and some night sweats. Pertinent negative results on physical examination
include the normal jugular venous pressure and
normal cardiac and lung findings. The absence
of wheezing, however, does not rule out obstructive lung disease a chest radiograph should be
obtained. Ambulatory oxygen saturation should
also be assessed, since patients with gas-exchange
abnormalities resulting from lung disease may
have desaturation with exercise, and such testing
provides information on the patients functional
capacity. In addition, the nature of changes in
the heart rate in response to exercise can be informative, since a rapid increase in the rate may
indicate poor cardiac function or arrhythmia.
A complete blood count (with a differential count)
and the results of serum chemical tests (including
tests of liver function and renal function) were
within normal limits. A chest radiograph revealed

n engl j med 364;1 nejm.org january 6, 2011

The New England Journal of Medicine

Downloaded from nejm.org on July 20, 2011. For personal use only. No other uses without permission.
Copyright 2011 Massachusetts Medical Society. All rights reserved.

73

The

n e w e ng l a n d j o u r na l

m e dic i n e

the presence of extensive pulmonary opacities,

with the most severe in the left perihilar region,
and diffuse bilateral involvement sparing the apexes (Fig. 1A and 1B). A 5-day course of azithromycin was prescribed, with no radiographic or symptomatic improvement.

Given the diffuse parenchymal radiographic abnormalities, a repeat course of antibiotics is appropriate. The persistence of symptoms after
multiple courses of antibiotics suggests a cause
other than community-acquired pneumonia and
demands a broadened differential diagnosis and
more aggressive evaluation. The laboratory data
and physical examination do not provide evidence
of extrapulmonary or systemic disease. The absence of an eosinophilia reduces the likelihood
of an asthma-plus syndrome.
Sputum smears should be obtained for Grams
staining, staining for acid-fast bacilli, and fungal
culture. An assessment of immune status is also
warranted and should include an HIV test and
measurement of antinuclear antibodies and serum immunoglobulin levels. Tests for hypersensitivity pneumonitis and for sensitivities to mold
mixes and bird-feather antigens should be performed, as well as a test for antineutrophil cytoplasmic antibodies (ANCA). In addition, pulmonary-function tests and a computed tomographic
(CT) scan of the chest are needed for a more
complete characterization of the patients diffuse
parenchymal lung disease.

Figure 1. Imaging Studies of the Chest.

Radiographs of the chest (posteroanterior view in Panel
A and lateral view in Panel B) show bilateral, predominantly perihilar, hazy air-space opacities. The pulmonary
vasculature appears normal. A computed tomographic
scan of the chest (Panel C) shows patchy ground-glass
opacities with associated intralobular septal thickening
(crazy paving, arrow), without involvement of the
interlobular septa.

74

of

n engl j med 364;1

A sputum culture grew oral flora. Pulmonaryfunction testing showed a forced vital capacity
(FVC) of 1.85 liters (58% of the predicted value),
a forced expiratory volume in 1 second (FEV1) of
1.8 liters (64% of the predicted value), and a ratio
of FEV1 to FVC of 97%. Total lung capacity was
1.82 liters (41% of the predicted value), and residual volume was 0.46 liters (41% of the predicted
value). The corrected carbon monoxide diffusing
capacity was 36% of the predicted value. Treadmill
testing showed a baseline oxygen saturation of
97%, with desaturation to 85% while the patient
was walking.
The pulmonary-function tests point to a restrictive ventilatory defect, a finding that is consistent
with the diffuse air-space disease observed on
radiography. Measurements of lung volume are
also important for the assessment of airflow obstruction. In patients with restriction, a low-value
nejm.org

january 6, 2011

The New England Journal of Medicine

Downloaded from nejm.org on July 20, 2011. For personal use only. No other uses without permission.
Copyright 2011 Massachusetts Medical Society. All rights reserved.

clinical problem-solving

FVC can mask the presence of an obstruction on

spirometry by bringing the ratio of FEV1 to FVC
into the normal range. On lung-volume testing,
however, airflow obstruction and the resulting
trapped gas may be evident when there is a relative or absolute elevation of the residual volume.
The observation of a low residual volume in this
case makes asthma or an asthma-plus syndrome
unlikely. The low saturation of ambulatory oxygen reflects compromised gas exchange.
The serum level of angiotensin-converting enzyme
was 23 U per liter (normal range, 9 to 67), and the
lactate dehydrogenase level was 153 U per liter
(normal range, 100 to 250). An antinuclear-antibody screen was negative, as was the result of a
tuberculin skin test. A CT scan of the chest revealed extensive bilateral ground-glass opacities
and septal thickening without apparent hilar
adenopathy (Fig. 1C).
Although an elevated level of angiotensin-converting enzyme is present in the majority of patients with sarcoidosis, the tests sensitivity and
specificity are insufficient to warrant its inclusion
in this diagnostic workup. Similarly, measurement of the level of lactate dehydrogenase does
not help to refine the differential diagnosis. The
bilateral ground-glass opacities seen on the chest
CT scan represent alveolar filling, which points
to a differential diagnosis that includes diffuse
alveolar hemorrhage, infection, edema, cancer
(e.g., bronchoalveolar carcinoma or primary pulmonary lymphoma), and interstitial lung disease,
a category encompassing diffuse parenchymal
lung processes that often involve both the interstitium and the alveolar space. In this case,
hemorrhage, edema, and infection are unlikely,
although we cannot rule out atypical chronic
infections, such as tuberculosis or pneumocystis
pneumonia. A tumor would also be unlikely, given the patients age and the clinical progression.
Interstitial lung diseases include conditions
resulting from occupational or environmental
exposures (e.g., hypersensitivity pneumonitis),
drug-induced lung injury, pneumonitis associated with collagen vascular disease, inherited processes (e.g., tuberous sclerosis), and other specific syndromes, such as sarcoidosis, alveolar
proteinosis, the ChurgStrauss syndrome, and
chronic eosinophilic pneumonia. Finally, there
are the remaining idiopathic interstitial pneumonias, which infrequently have pathognomonic

radiographic or clinical findings. Our patient is

a nonsmoker who has not had obvious environmental or toxic exposures likely to cause hypersensitivity pneumonitis or lung injury, although
some exposures can be difficult to identify. The
possibility of sicca-type symptoms notwithstanding, there is no evidence of a collagen vascular
disease, and a test for antinuclear antibodies was
negative.
In addition to showing ground-glass opaci
ties, the CT images reveal extensive lobular septal
thickening, a diagnostic turning point in this
case. Although septal thickening often represents
nonspecific accumulation of interstitial fluid
most commonly, pulmonary edema the combination of septal thickening and variable alveolar filling can be manifested as a crazy paving
radiographic pattern. Crazy paving is highly
characteristic of pulmonary alveolar proteinosis,
although it can also be observed in other diseases (e.g., alveolar sarcoidosis, lipoid pneumonia, and mucinous bronchoalveolar carcinoma).
The pathophysiological basis of pulmonary alveolar proteinosis is primary macrophage dysfunction, which allows surfactant and lipid-laden
macrophages to accumulate in the intraalveolar
and interstitial spaces. The CT scan should be
reviewed with a chest radiologist to determine
whether the noted septal thickening is consistent with crazy paving.
A review of the chest CT showed that the septal
thickening had a pattern consistent with crazy paving, predominately in the lower lobes and perihilar regions. This regional distribution is frequently observed in pulmonary alveolar proteinosis.
Bronchoscopy with bronchoalveolar lavage and
transbronchial lung biopsy constitute the preferred diagnostic procedure for pulmonary alveolar proteinosis, but a surgical lung biopsy is also
reasonable. If this diagnosis is confirmed by histologic analysis, the level of granulocytemacrophage colony-stimulating factor (GM-CSF) antibodies should be measured to determine whether
this is a case of autoimmune pulmonary alveolar
proteinosis. Treatment of autoimmune pulmonary
alveolar proteinosis focuses on symptom management. Treatment options include whole-lung
lavage, GM-CSF administration, immunosuppression (with rituximab or mycophenolate mofetil),
or a combination of these treatments. In contrast, other cases of pulmonary alveolar pro-

n engl j med 364;1 nejm.org january 6, 2011

The New England Journal of Medicine

Downloaded from nejm.org on July 20, 2011. For personal use only. No other uses without permission.
Copyright 2011 Massachusetts Medical Society. All rights reserved.

75

The

n e w e ng l a n d j o u r na l

m e dic i n e

teinosis (those in which GM-CSF antibodies are

undetectable) are considered to be a result of
other disease processes that impair the function
of alveolar macrophages (e.g., hematologic cancer, immunosuppression, or toxic inhalation), and
treatment of these cases is directed at the underlying condition or cause.

A wedge-biopsy specimen of the left upper lobe,

obtained by means of video-assisted thoracoscopic surgery, revealed alveolar phospholipoproteinosis (Fig. 2). The level of antiGM-CSF antibodies
was elevated, at a titer of 1:12,800. Multiple lung
lavages were performed and provided some symptomatic improvement, but the treatment had to
be repeated up to four times a year for symptom
control. Subsequent trials of rituximab and mycophenolate mofetil were without benefit. For the
past 3 years, the patient has been treated with inhaled GM-CSF twice daily. She has had a good
clinical response and has not required whole-lung
lavage since beginning this medication. She is
able to do office work and exercises at low intensity on a regular basis.

C om men ta r y

Figure 2. Lingular-Biopsy Specimens.

The lingular-biopsy tissue shows proteinaceous material
filling alveoli (Panel A, hematoxylin and eosin), a finding
that is characteristic of pulmonary alveolar phospholipoproteinosis; prominent lobular septa correlate with
the crazy paving seen on radiography. The intraalveolar
material is palely eosinophilic, granular, and largely
acellular (Panel B, hematoxylin and eosin). Periodic
acidSchiff staining (Panel C) reflects a composition
that is rich in pulmonary surfactant (deep magenta).

76

of

n engl j med 364;1

The patients symptoms cough and shortness

of breath are seen often in primary care. In
the majority of cases, the underlying process is
benign. How, then, do physicians discriminate
between patients requiring an aggressive and expensive evaluation and those whose care can be
managed conservatively?
In adults, the differential diagnosis of and
clinical approach to cough are guided by the
duration of symptoms, from acute (less than
3 weeks) to subacute (3 to 8 weeks) to chronic
(more than 8 weeks).1 This patient presented
with a history of cough and shortness of breath.
Although chronic cough is most often caused by
postnasal drip, gastroesophageal reflux, or asthma, the progression of symptoms and the lack of
response to empirical asthma therapy in this case
pointed to causes that are less common and
more ominous.
The shortness of breath, in particular, warranted an aggressive diagnostic evaluation, including ambulatory oximetry and radiographic
assessment. Measurement of ambulatory oxygen
saturation at the patients second presentation
would probably have facilitated an earlier diag-

nejm.org

january 6, 2011

The New England Journal of Medicine

Downloaded from nejm.org on July 20, 2011. For personal use only. No other uses without permission.
Copyright 2011 Massachusetts Medical Society. All rights reserved.

clinical problem-solving

nosis. Instead, trials of empirical therapy were

selected to address common causes of cough
that did not readily explain her associated (and
progressive) shortness of breath. The patients
history of asthma probably gave her physicians
a false sense of assurance that asthma was the
cause of her symptoms, when in fact her presentation and course were atypical for asthma.
Pulmonary alveolar proteinosis is a rare disease, with a reported incidence of 1 case per
300,000 persons in the United States.2,3 The
disease is more common in men than in women
(occurring at a ratio of approximately 4 to 1),
and the median age at diagnosis is 39 years.2-4
Therefore, this case reflects a somewhat atypical
presentation of a rare disease that was nonetheless suspected once the characteristic crazy-paving pattern5 was noted on imaging. Pulmonary
alveolar proteinosis often presents initially as an
apparent bilateral, community-acquired pneumonia, which then fails to resolve despite multiple
courses of appropriate antibiotics. Radiographic
findings are often more impressive than report
ed symptoms, a clinical discordance that can help
direct the diagnostic evaluation. The disease can
be congenital but is more commonly acquired.2-4
Congenital pulmonary-surfactant derangements
are caused by mutations in surfactant protein
genes or GM-CSFreceptor genes6; these derangements vary in clinical severity and are increasingly recognized as a cause of adult interstitial
lung disease with a radiographic pattern similar to that of pulmonary alveolar proteinosis.
Acquired pulmonary alveolar proteinosis can
result when alveolar macrophage function is
References
1. Irwin RS, Baumann MH, Bolser DC,
et al. Diagnosis and management of
cough executive summary: ACCP evidence-based clinical practice guidelines.
Chest 2006;129:Suppl:1S-23S.
2. Trapnell BC, Whitsett JA, Nakata K.
Pulmonary alveolar proteinosis. N Engl J
Med 2003;349:2527-39.
3. Ioachimescu OC, Kavuru MS. Pulmonary alveolar proteinosis. Chron Respir
Dis 2006;3:149-59.
4. Seymour JF, Presneill JJ. Pulmonary
alveolar proteinosis: progress in the first

impaired or reduced because of inhalational exposures (e.g., exposure to silica or insecticides),

hematologic cancers, or immunosuppression
(e.g., as a result of HIV infection or immunosuppressive medications); symptoms generally resolve with treatment of the underlying cause.
However, the majority of acquired cases (approximately 90%) result from an autoimmune
process with neutralizing antiGM-CSF antibodies (with evidence of causality in a mouse model7), not a GM-CSF deficiency.
In case series, GM-CSF administration has
been associated with improvement in oxygena
tion and quality of life in about 50% of acquired
cases8; randomized trials are needed to further
assess the efficacy of this therapy. In a retrospective analysis, the 10-year rate of survival
from pulmonary alveolar proteinosis increased
from 85% among patients who were not treated
with lavage to 95% in patients who were treated
with lavage, with up to 85% of patients having
immediate improvement in symptoms, oxygenation, or radiographic signs of disease.4 At present, whole-lung lavage remains the treatment of
choice. Patients requiring repeated alveolar lavages have a poorer prognosis and greater rate
of progression to fibrosis. Lung transplantation
is a therapeutic option for refractory pulmonary
alveolar proteinosis, but its long-term benefit is
not well characterized; the disease can recur in
the lung allograft.9
Financial and other disclosures provided by the authors are
available with the full text of this article at NEJM.org.
We thank Sara Vargas for her assistance with the histopathological analysis and Susan Vignolo-Collazzo for assistance with
preparation of an earlier version of this article.

44 years. Am J Respir Crit Care Med

2002;166:215-35.
5. Frazier AA, Franks TJ, Cooke EO, Mohammed T-LH, Pugatch RD, Galvin JR.
From the archives of the AFIP: pulmonary
alveolar proteinosis. Radiographics 2008;
28:883-99.
6. Suzuki T, Sakagami T, Rubin BK, et al.
Familial pulmonary alveolar proteinosis
caused by mutations in CSF2RA. J Exp
Med 2008;205:2703-10.
7. Uchida K, Beck DC, Yamamoto T,
et al. GM-CSF autoantibodies and neutro-

phil dysfunction in pulmonary alveolar

proteinosis. N Engl J Med 2007;356:56779.
8. Venkateshiah SB, Yan TD, Bonfield TL,
et al. An open-label trial of granulocyte
macrophage colony stimulating factor
therapy for moderate symptomatic pulmonary alveolar proteinosis. Chest 2006;
130:227-37.
9. Parker LA, Novotny DB. Recurrent alveolar proteinosis following double lung
transplantation. Chest 1997;111:1457-8.
Copyright 2011 Massachusetts Medical Society.

n engl j med 364;1 nejm.org january 6, 2011

The New England Journal of Medicine

Downloaded from nejm.org on July 20, 2011. For personal use only. No other uses without permission.
Copyright 2011 Massachusetts Medical Society. All rights reserved.

77