Practice Essentials
Pulmonary alveolar proteinosis (PAP) is a rare, diffuse lung disease that is characterized by the alveolar and interstitial accumulation of a periodic acid-Schiff (PAS) stain-positive phospholipoprotein that is derived from surfactant. The lung architecture is otherwise normal, and any associated inflammation or fibrosis is limited in extent. High-resolution CT (HRCT) scanning is superior to conventional CT scanning and to chest radiography for demonstrating the morphologic characteristics of PAP. Crazy paving (see the second image below) is the characteristic finding in PAP on HRCT scanning; it consists of patchy, bilateral geographic areas of ground-glass opacity, which are associated with interlobular septal thickening.[1, 2, 3, 4, 5, 6, 7, 8]
There are 3 types of PAP: primary, secondary, and congenital. The most frequent form is primary PAP, which includes autoimmune PAP; it is defined by the presence of circulating anti-GM-CSF antibodies. Secondary PAP is primarily caused by hematologic disease, infections, or inhaling toxic substances. Congenital PAP affects children almost exclusively.[1]
(A radiograph and a CT scan of PAP are shoiwn below.)
Frontal chest radiograph from a patient with pulmonary alveolar proteinosis. This image demonstrates bilateral perihilar and infrahilar ground-glass opacity without evidence of mediastinal widening, pleural effusion, or adenopathy.
High-resolution computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) in a patient with acute silicoproteinosis. This image demonstrates crazy paving, which refers to bilateral ground-glass opacity that is associated with marked interlobular septal thickening and a sharp nonanatomic demarcation between normal and abnormal lung.
Primary PAP is caused by disruption of granulocyte-macrophage colony-stimulating factor (GM-CSF) signaling, either by GM-CSF autoantibodies (ie, autoimmune PAP)or genetic mutations involving the GM-CSF receptor (ie, hereditary PAP). Secondary PAP occurs in a heterogeneous group of conditions that reduce numbers or functions of alveolar macrophages and thereby surfactant clearance. Congenital PAP is caused by mutations in genes required for normal surfactant production. Over 90% of cases are autoimmune. Hereditary PAP accounts for approximately 3% of cases, secondary PAP for 4%, and congenital PAP for 1.5%.[9, 10]
Peferred examination
Although CT scan (in particular, high-resolution CT [HRCT] scan) findings are often characteristic of PAP, fiberoptic bronchoscopy with bronchoalveolar lavage and transbronchial biopsy are sufficient for confirming the clinical diagnosis in most patients. Chest radiographs alone, while occasionally suggestive, are rarely diagnostic in patients with PAP. CT scanning, especially HRCT scanning, often displays findings that are characteristic of, although not pathognomonic for, PAP.[11, 12, 13, 14, 15, 16, 5] Indeed, Johkoh et al include at least 15 different entities in the differential diagnosis for the characteristic HRCT scan findings of PAP.[17]
Few data are available regarding the MRI appearance of pulmonary alveolar proteinosis (PAP). Because of magnetic susceptibility effects and the short echo times of lung parenchyma, lung tissue is challenging to image with MRI.
A study by Moore et al addressed the appearances of diffuse air-space disease on MRI and found that air-space opacity in PAP has a short T1 value. This opacity demonstrated relatively little increased signal intensity with T2-weighted imaging.[18] Presumably, the short T1 signal of PAP reflects the relative lack of water within the lipoproteinaceous material that fills the air spaces in this disease.
A serum GM-CSF autoantibody test is highly sensitive and specific for autoimmune PAP. A negative GM-CSF autoantibody test excludes autoimmune PAP but does not exclude other PAP-causing diseases. However, additional blood-based tests to measure GM-CSF receptors/signaling and other genetic tests are available and permit the rapid, accurate diagnosis of the PAP-causing disease in more than 95% of patients. Because lung biopsy is associated with significant procedure-related morbidity, it should be reserved for patients in whom identification of the cause of PAP remains uncertain after completion of disease-specific, noninvasive serologic, blood-based, and genetic tests.[19]
Limitations of techniques
The differential diagnosis of PAP syndrome is challenged by its low prevalence, nonspecific symptoms (dyspnea, cough, chest tightness, and fatigue), common radiographic appearance (diffuse, patchy, ground-glass opacification and septal thickening), and normal routine laboratory tests. Many patients with PAP undergo transbronchial biopsy, surgical lung biopsy, or both, and histologic examination fails to identify the presence of PAP in 28% of cases because of patchy involvement. PAP is often misdiagnosed as pneumonia initially until the failure of multiple courses of empirical antibiotic therapy prompts diagnostic reconsideration and referral, thereby delaying accurate diagnosis by a median of 1.5 years (range, 0.5-13.5 years).[19]
Radiography
The classic chest radiographic finding of pulmonary alveolar proteinosis (PAP) is bilateral, symmetrical air-space opacity, which appears either as ground-glass or frank consolidation, with a perihilar or basilar predominance (see the images below). Air bronchograms are uncommon.
Frontal chest radiograph in a patient with pulmonary alveolar proteinosis. This image reveals bilateral air-space opacity without evidence of effusion or mediastinal widening. A faintly reticular pattern is present, representing thickened, interlobular septa.
Frontal chest radiograph in a patient with pulmonary alveolar proteinosis that was subsequently treated with bronchoalveolar lavage. This image reveals bilateral, symmetrical air-space opacity without pleural effusion or mediastinal widening.
Chest radiograph in a patient after bronchoalveolar lavage. This image demonstrates improvement in the bilateral air-space opacity. These opacities subsequently cleared completely, and the patient remained in remission.
Frontal chest radiograph in a patient with the diffuse form of bronchioloalveolar carcinoma. This image shows the presence of bilateral, hazy, ground-glass attenuation; a normal mediastinal width; and no pleural effusion. This appearance resembles alveolar proteinosis.
Radiographic opacities are often vaguely nodular and may be accompanied by fine, linear opacities or reticulation. Serial radiographs may demonstrate persistence of this pattern over time, resulting in a more limited differential diagnosis of chronic air-space opacity.
Unlike in hydrostatic pulmonary edema, the mediastinum is not widened, the heart is not enlarged, and pleural effusions are not common. Adenopathy is not a feature of the disease.
Atypical manifestations include a linear or reticular pattern that is unaccompanied by air-space opacity, lower lobe predominant consolidation, or poorly defined nodules as the primary radiographic manifestation of disease.
The differential diagnosis of PAP on chest radiographs includes all entities that result in diffuse air-space disease, such as hydrostatic and noncardiogenic pulmonary edema, diffuse infections (particularly Pneumocystis pneumonia), pulmonary hemorrhage, interstitial pneumonia, the diffuse form of bronchioloalveolar carcinoma, toxic lung injuries and hypersensitivity pneumonitis, eosinophilic pneumonia, drug or radiation-induced pulmonary disease, and lipoid pneumonia.
Chest radiograph and CT scan findings of treated PAP have also been described.[20, 21, 22]
The initial radiographs after BAL reveal increased opacity in the lobes in which the BAL fluid was instilled; this opacity is transient, clearing in hours to a few days.
After BAL, the aeration of the washed lung gradually improves. By 6 weeks after the BAL procedure, chest radiographs usually demonstrate improved aeration over preprocedure radiographs.
Rarely, in the course of resolution, new opacities may develop in either the washed lung or the contralateral lung. Occasionally, these opacities are related to endobronchial obstruction, usually representing atelectasis. Areas of decreased lung opacity, which represent hyperinflation, may also occur by the same mechanism.
Opacity in the contralateral lung may result from spillage of BAL fluid from the washed lung or atelectasis after the BAL procedure.
Complications of the BAL procedure, including pneumomediastinum and pneumothorax, may be evident on immediate postprocedure radiographs.
Degree of confidence
Radiographic findings of PAP are not specific, and as stated above, a large differential diagnosis must be considered. Similar to the case with other diffuse lung diseases, HRCT scanning provides a more specific morphologic evaluation of the disease pattern and may provide a more limited differential diagnosis.
Computed Tomography
Routine CT scans (7-10 mm collimation) may reveal bilateral areas of consolidation and reticulation in the lungs of patients with pulmonary alveolar proteinosis (PAP) (see the images below).[11, 14, 15, 23, 24, 25] The disease distribution of PAP is usually bilateral and patchy.
Occasionally, an underlying linear abnormality that represents thickened interlobular septa may be appreciated.
HRCT scanning is superior to conventional CT scanning and to chest radiography for demonstrating the morphologic characteristics of PAP.[26, 27] Crazy paving is the characteristic finding in PAP on HRCT scanning; it consists of patchy, bilateral geographic areas of ground-glass opacity, which are associated with interlobular septal thickening. The HRCT scan features of PAP are seen in the images below.[12, 16, 17, 5, 7]
High-resolution computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) in a patient with pulmonary alveolar proteinosis. The image reveals bilateral ground-glass opacity that is associated with septal thickening, which is consistent with the crazy-paving pattern.
High-resolution computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) in a patient with pulmonary alveolar proteinosis who received bronchoalveolar lavage. This image reveals regression of the ground-glass opacity, representing partial resolution of the disease.
High-resolution computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) in a patient with pulmonary alveolar proteinosis. This image demonstrates the presence of bilateral ground-glass opacity and interlobular septal thickening.
High-resolution computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) in a patient with pulmonary alveolar proteinosis (PAP) (same patient as in the previous image). This image reveals slight interval worsening of the bilateral ground-glass opacity and interlobular septal thickening that is consistent with progression of PAP.
PAP is often distributed uniformly from the lung apex to the base. Interlobular septal thickening may be encountered more frequently in the lower lung zones. Although the ground-glass opacity in PAP is usually patchy, centrilobular nodules have been described in pediatric patients with PAP, as well as in adult patients with acute silicoproteinosis.[13]
Although findings of the interlobular septa are commonly abnormal for pathologic specimens from patients with PAP, septa findings may occasionally be pathologically normal, despite a thickened appearance on HRCT scans. This apparent septal thickening on HRCT scans presumably reflects the aggregation of PAS-positive lipoproteinaceous material immediately adjacent to the interlobular septa.
Classically, the abnormal pulmonary parenchyma is demarcated sharply from normal lung areas without a discernible anatomic boundary. (The images below show no such sharp demarcation and, therefore, suggest a diagnosis other than PAP.)
Axial computed tomography scan (window width = 1500 Hounsfield units [HU], level = -600 HU) in a patient with the diffuse form of bronchioloalveolar carcinoma. Note the presence of ground-glass opacity that is associated with septal thickening, resembling the crazy-paving appearance of alveolar proteinosis. The space-occupying nature of the process (note that the left major fissure is bowed posteriorly) and the lack of the usually sharp, but nonanatomic, demarcation between the normal and abnormal lung (best seen in the posterior right lung) are findings that are not characteristic of alveolar proteinosis and, thus, should suggest an alternative diagnosis.
Axial computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) in a patient with diffuse pulmonary hemorrhage. This image reveals diffuse, bilateral, ground-glass opacity that is associated with mild, interlobular septal thickening. The lack of the sharp demarcation between the normal and abnormal pulmonary parenchyma, which is characteristic of crazy paving, suggests a diagnosis other than alveolar proteinosis. In addition, the septal thickening in alveolar proteinosis is usually more prominent.
Axial computed tomography scan (window width = 1000 Hounsfield units [HU], level = -700 HU) obtained through the lower lobes in a patient with pulmonary edema. This image reveals the bilateral, ground-glass opacity that is associated with smooth, interlobular, septal thickening. Although by definition these findings resemble crazy paving, the appearance is nevertheless quite distinct from alveolar proteinosis. Note also the absence of the sharp demarcation between the normal and abnormal lung that is usually a characteristic of crazy paving in alveolar proteinosis.
Similar to the case with other diffuse lung diseases, HRCT scanning provides a more specific morphologic evaluation of the disease pattern and may provide a more limited differential diagnosis. HRCT scanning is also useful for targeting the optimal site for tissue sampling.
Generally, PAP in children is often more acute and severe than PAP in adults. Chest radiographic and HRCT scan findings in childhood forms of PAP are few but are likely similar to their adult counterparts.
McCook et al suggested that a miliarylike pattern is encountered more frequently in the pediatric entity than in the adult form and that basilar predominant linear and reticular abnormalities may be observed more often on CT scans in children.[28]
HRCT scan descriptions of childhood PAP by Zontsich et al suggest that the same crazy-paving pattern seen in adults may be seen in children.[21]
Pleural effusions and adenopathy are uncommon in PAP, and the clinician should suspect a superimposed infection or malignancy if these are present. Occasionally, pleural effusions may be encountered in uncomplicated cases of PAP shortly after BAL.
Degree of confidence
Although the appearance of crazy paving is highly suggestive of PAP, this finding is not pathognomonic for PAP. Many other entities may present with a pattern similar to crazy paving on HRCT scans, including the following:
-
Hydrostatic pulmonary edema
-
Diffuse alveolar damage from any number of causes
-
-
Diffuse pulmonary infections (including Mycobacterium tuberculosis, M pneumoniae, other bacterial pneumonias)
-
Diffuse form of bronchioloalveolar carcinoma
-
Acute respiratory distress syndrome (ARDS)
-
-
-
Bronchiolitis obliterans organizing pneumonia
-
Chronic eosinophilic pneumonia
-
-
Acute interstitial pneumonia
-
Atypical manifestations of PAP also occur. Do not expect all instances of PAP to present as crazy paving on HRCT scans.
Author
Michael B Gotway, MD Professor, Department of Radiology, Mayo Clinic Scottsdale; Clinical Associate Professor, Diagnostic Radiology/Biomedical Imaging and Pulmonary/Critical Care Medicine, Department of Radiology, University of California, San Francisco, School of Medicine; Clinical Professor, University of Arizona College of Medicine; Adjunct Professor, Department of Biomedical Informatics, Arizona State University
Michael B Gotway, MD is a member of the following medical societies: American College of Radiology, American Roentgen Ray Society, Society of Thoracic Radiology, North American Society for Cardiovascular Imaging, Radiological Society of North America
Disclosure: Nothing to disclose.
Coauthor(s)
Theodore J Lee, MD Chief of Thoracic Radiology, San Francisco General Hospital; Assistant Professor of Clinical Radiology, University of California, San Francisco, School of Medicine
Theodore J Lee, MD is a member of the following medical societies: American Roentgen Ray Society, Radiological Society of North America, Society of Thoracic Radiology
Disclosure: Nothing to disclose.
Specialty Editor Board
Bernard D Coombs, MB, ChB, PhD Consulting Staff, Department of Specialist Rehabilitation Services, Hutt Valley District Health Board, New Zealand
Disclosure: Nothing to disclose.
Chief Editor
Eugene C Lin, MD Attending Radiologist, Teaching Coordinator for Cardiac Imaging, Radiology Residency Program, Virginia Mason Medical Center; Clinical Assistant Professor of Radiology, University of Washington School of Medicine
Eugene C Lin, MD is a member of the following medical societies: American College of Nuclear Medicine, American College of Radiology, Radiological Society of North America, Society of Nuclear Medicine and Molecular Imaging
Disclosure: Nothing to disclose.
Additional Contributors
Jeffrey A Miller, MD Associate Adjunct Professor of Clinical Radiology, University of Medicine and Dentistry of New Jersey-New Jersey Medical School; Faculty, Department of Radiology, Veterans Affairs of New Jersey Health Care System
Jeffrey A Miller, MD is a member of the following medical societies: American Roentgen Ray Society, Radiology Alliance for Health Services Research, Society of Thoracic Radiology
Disclosure: Nothing to disclose.
Acknowledgements
The authors and editors of Medscape Reference gratefully acknowledge the contributions of previous author Samuel K Dawn, MD, to the development and writing of this article.
Jouneau S, Ménard C, Lederlin M. Pulmonary alveolar proteinosis. Respirology. 2020 May 3. [QxMD MEDLINE Link].
Mlika M, Basit H, Kaul P. Alveolar Proteinosis. 2020 Jan. [QxMD MEDLINE Link]. [Full Text].
Trapnell BC, Nakata K, Bonella F, Campo I, Griese M, Hamilton J, et al. Pulmonary alveolar proteinosis. Nat Rev Dis Primers. 2019 Mar 7. 5 (1):16. [QxMD MEDLINE Link].
Rajaonarison Ny Ony Narindra LH, Andrianah EG, Ranaivomanana VF, Tomboravo C, Ranoharison HD, Bruneton JN, et al. Atypical alveolar proteinosis. Indian J Radiol Imaging. 2018 Oct-Dec. 28 (4):439-441. [QxMD MEDLINE Link].
Da Nam B, Kim TJ, Chung MP, Chung MJ, Kim TS, Lee KS. CT findings in pulmonary alveolar proteinosis: serial changes and prognostic implications. J Thorac Dis. 2018 Oct. 10 (10):5774-5783. [QxMD MEDLINE Link].
Kiani A, Parsa T, Adimi Naghan P, Dutau H, Razavi F, Farzanegan B, et al. An eleven-year retrospective cross-sectional study on pulmonary alveolar proteinosis. Adv Respir Med. 2018. 86 (1):7-12. [QxMD MEDLINE Link].
Matsuura H, Yamaji Y. Pulmonary Alveolar Proteinosis: Crazing-Paving Appearance. Am J Med. 2018 Apr. 131 (4):e153-e154. [QxMD MEDLINE Link].
Griese M. Pulmonary Alveolar Proteinosis: A Comprehensive Clinical Perspective. Pediatrics. 2017 Aug. 140 (2):[QxMD MEDLINE Link].
McCarthy C, Avetisyan R, Carey BC, Chalk C, Trapnell BC. Prevalence and healthcare burden of pulmonary alveolar proteinosis. Orphanet J Rare Dis. 2018 Jul 31. 13 (1):129. [QxMD MEDLINE Link]. [Full Text].
Kitamura N, Ohkouchi S, Tazawa R, Ishii H, Takada T, Sakagami T, et al. Incidence of autoimmune pulmonary alveolar proteinosis estimated using Poisson distribution. ERJ Open Res. 2019 Feb. 5 (1):[QxMD MEDLINE Link]. [Full Text].
Guan Y, Zeng Q, Yang H, Zheng J, Li S, Gao Y, et al. Pulmonary alveolar proteinosis: quantitative CT and pulmonary functional correlations. Eur J Radiol. 2012 Sep. 81(9):2430-5. [QxMD MEDLINE Link].
Rossi SE, Erasmus JJ, Volpacchio M, et al. "Crazy-paving" pattern at thin-section CT of the lungs: radiologic-pathologic overview. Radiographics. 2003 Nov-Dec. 23(6):1509-19. [QxMD MEDLINE Link]. [Full Text].
Marchiori E, Ferreira A, Muller NL. Silicoproteinosis: high-resolution CT and histologic findings. J Thorac Imaging. 2001 Apr. 16(2):127-9. [QxMD MEDLINE Link].
Wang YL, Fang N, Zeng L, Wu ZJ, Cui XJ. Localized Airspace Consolidation of Pulmonary Alveolar Proteinosis Mimicking Malignant Lesions in 18F-FDG PET/CT Imaging: One Case Report. Clin Nucl Med. 2015 Nov. 40 (11):908-9. [QxMD MEDLINE Link].
Danilevskaya O, Averyanov A, Lesnyak V, Chernyaev A, Sorokina A. Confocal laser endomicroscopy for diagnosis and monitoring of pulmonary alveolar proteinosis. J Bronchology Interv Pulmonol. 2015 Jan. 22 (1):33-40. [QxMD MEDLINE Link].
Luo J, Yang D, Zhou S, Xiao E, Chen P, Fan S, et al. Crazy paving pattern caused by pulmonary alveolar proteinosis: CT findings and the pathologic basis. Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2014 Sep. 39 (9):924-9. [QxMD MEDLINE Link].
Johkoh T, Itoh H, Muller NL, et al. Crazy-paving appearance at thin-section CT: spectrum of disease and pathologic findings. Radiology. 1999 Apr. 211(1):155-60. [QxMD MEDLINE Link]. [Full Text].
Moore EH, Webb WR, Muller N, Sollitto R. MRI of pulmonary airspace disease: experimental model and preliminary clinical results. AJR Am J Roentgenol. 1986 Jun. 146(6):1123-8. [QxMD MEDLINE Link].
McCarthy C, Carey B, Trapnell BC. Blood Testing for Differential Diagnosis of Pulmonary Alveolar Proteinosis Syndrome. Chest. 2019 Feb. 155 (2):450-452. [QxMD MEDLINE Link]. [Full Text].
Gale ME, Karlinsky JB, Robins AG. Bronchopulmonary lavage in pulmonary alveolar proteinosis: chest radiograph observations. AJR Am J Roentgenol. 1986 May. 146(5):981-5. [QxMD MEDLINE Link]. [Full Text].
Zontsich T, Helbich TH, Wojnarovsky C, et al. Pulmonary alveolar proteinosis in a child: HRCT findings before and after bronchoalveolar lavage. Eur Radiol. 1998. 8(9):1680-2. [QxMD MEDLINE Link].
Stoica R, Macri A, Cordos I, Bolca C. Whole lung lavage for pulmonary alveolar proteinosis after surgery for spontaneous pneumothorax. J Med Life. 2012 Sep 15. 5(3):311-5. [QxMD MEDLINE Link]. [Full Text].
Godwin JD, Muller NL, Takasugi JE. Pulmonary alveolar proteinosis: CT findings. Radiology. 1988 Dec. 169(3):609-13. [QxMD MEDLINE Link].
Holbert JM, Costello P, Li W, et al. CT features of pulmonary alveolar proteinosis. AJR Am J Roentgenol. 2001 May. 176(5):1287-94. [QxMD MEDLINE Link].
Murch CR, Carr DH. Computed tomography appearances of pulmonary alveolar proteinosis. Clin Radiol. 1989 May. 40(3):240-3. [QxMD MEDLINE Link].
Lee KN, Levin DL, Webb WR, et al. Pulmonary alveolar proteinosis: high-resolution CT, chest radiographic, and functional correlations. Chest. 1997 Apr. 111(4):989-95. [QxMD MEDLINE Link].
Santamaria F, Parenti G, Guidi G, et al. Early detection of lung involvement in lysinuric protein intolerance: role of high-resolution computed tomography and radioisotopic methods. Am J Respir Crit Care Med. 1996 Feb. 153(2):731-5. [QxMD MEDLINE Link].
McCook TA, Kirks DR, Merten DF, et al. Pulmonary alveolar proteinosis in children. AJR Am J Roentgenol. 1981 Nov. 137(5):1023-7. [QxMD MEDLINE Link].
Brasch F, Birzele J, Ochs M, et al. Surfactant proteins in pulmonary alveolar proteinosis in adults. Eur Respir J. 2004 Sep. 24(3):426-35. [QxMD MEDLINE Link]. [Full Text].
Seymour JF, Presneill JJ. Pulmonary alveolar proteinosis: Progress in the first 44 years. Am J Respir Crit Care Med. July 2002. 166:215-35. [QxMD MEDLINE Link]. [Full Text].
Shah PL, Hansell D, Lawson PR, et al. Pulmonary alveolar proteinosis: clinical aspects and current concepts on pathogenesis. Thorax. 2000 Jan. 55(1):67-77. [QxMD MEDLINE Link]. [Full Text].
Carbonelli C, Prati F, Carretto E, Cavazza A, Spaggiari L, Magnani G. Cavitating pulmonary nodules growing in a favourable medium. Thorax. 2013 May 18. [QxMD MEDLINE Link].