Granulomatous lung disease: an approach to the differential diagnosis.
Abstract: Context.--Granulomas are among the most commonly encountered abnormalities in pulmonary pathology and often pose a diagnostic challenge. Although most pathologists are aware of the need to exclude an infection in this setting, there is less familiarity with the specific histologic features that aid in the differential diagnosis.

Objective.--To review the differential diagnosis, suggest a practical diagnostic approach, and emphasize major diagnostically useful histologic features. This review is aimed at surgical pathologists who encounter granulomas in lung specimens.

Data Sources.--Pertinent recent and classic peer-reviewed literature retrieved from PubMed (US National Library of Medicine) and primary material from the institutions of both authors.

Conclusions.--Most granulomas in the lung are caused by mycobacterial or fungal infection. The diagnosis requires familiarity with the tissue reaction as well as with the morphologic features of the organisms, including appropriate interpretation of special stains. The major noninfectious causes of granulomatous lung disease are sarcoidosis, Wegener granulomatosis, hypersensitivity pneumonitis, hot tub lung, aspiration pneumonia, and talc granulomatosis.

(Arch Pathol Lab Med. 2010;134:667-690)
Article Type: Report
Subject: Granuloma (Development and progression)
Granuloma (Diagnosis)
Granuloma (Care and treatment)
Histology, Pathological (Usage)
Authors: Mukhopadhyay, Sanjay
Gal, Anthony A.
Pub Date: 05/01/2010
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2010 College of American Pathologists ISSN: 1543-2165
Issue: Date: May, 2010 Source Volume: 134 Source Issue: 5
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 230246554
Full Text: The most common sites where surgical pathologists encounter granulomas are the skin and subcutaneous tissues, lymph nodes, and lungs. (1) Most pathologists respond appropriately to this finding by ordering "special stains" to exclude an infection. However, the process of detecting and identifying an organism frequently proves quite challenging. Most pathologists dread interpreting the special stains, namely, Ziehl-Neelsen (ZN) and Grocott methenamine silver (GMS). The task is even more arduous if, as is commonly the case, an organism is not detected. The lack of identifiable organisms in tissues raises questions about the next appropriate step for the pathologist to pursue. Does one dust off a lung pathology textbook, seek extramural consultation, or merely issue a descriptive diagnosis? What other diagnoses should be considered? Do the black particles on the GMS stain represent fungi? If the results obtained with special stains are negative, is the diagnosis sarcoidosis or could it be anything else? These are some of the problems that plague surgical pathologists on a daily basis.

The following review attempts to address these issues and will hopefully ease the task of approaching the differential diagnosis of granulomatous lung diseases. Since clinical and radiographic information is often unavailable to pathologists at the time of diagnosis, our emphasis will be on the histologic features. The reader is also referred to other excellent reviews on the subject. (2-4)

DEFINITION AND TERMINOLOGY

A granuloma is a compact aggregate of histiocytes (macrophages). (5) The histiocytes in granulomas are often described as "epithelioid." Epithelioid histiocytes have indistinct cell borders and elongated, sole-shaped nuclei, as opposed to the well-defined cell borders and round, oval, or kidney bean-shaped nuclei of ordinary histiocytes. Aggregation of histiocytes is the minimum requirement of a granuloma, regardless of whether the lesion also contains necrosis, lymphocytes, plasma cells, or multinucleated giant cells.

APPROACH TO THE DIFFERENTIAL DIAGNOSIS OF GRANULOMATOUS LUNG DISEASE

The differential diagnosis of granulomatous lung disease is listed in Table 1. The following is our recommended step-by-step approach to pulmonary granulomas.

Step 1: Attempt to identify an organism.

Step 2: Look for histologic features of noninfectious granulomatous diseases (Table 2).

Step 3: If steps 1 and 2 do not yield a specific diagnosis, ensure that an adequate number of blocks have been stained with special stains and reexamine the special stains. If no organisms are found despite a thorough reexamination, issue a descriptive diagnosis including the type of granulomas (necrotizing, nonnecrotizing, or both) and the absence of identifiable organisms. Suggest a differential diagnosis in a comment.

Step 1: Identifying Organisms

Since infection is the most common cause of pulmonary granulomas, it is always important to carefully exclude an infectious etiology before diagnosing a noninfectious granulomatous lung disease.

Organisms in Pulmonary Granulomas: What to Expect.--The organisms most commonly found in granulomas of the lung are mycobacteria and fungi. (1) Mycobacterium tuberculosis is common in the developing world, while nontuberculous mycobacteria (NTM) are frequently cultured from granulomas in the United States. The most common fungal causes of pulmonary granulomas are Histoplasma, Cryptococcus, Coccidioides, and Blastomyces. Prevalence of these fungi varies by geographic region. In the United States, Histoplasma is primarily seen in the central and eastern states, Coccidioides is endemic in the Southwest, Cryptococcus is ubiquitous, and Blastomyces is infrequent. (6-8) Aspergillus rarely causes granulomatous lung disease. Other fungi (Chrysosporium and Sporothrix) and bacteria (Brucella and Burkholderia) may also rarely cause granulomatous lung disease.

Using the Tissue Reaction as a Clue.--Organisms must be sought in both necrotizing and nonnecrotizing granulomas since they may be found in either type. (2,9) However, the search must be especially thorough in necrotizing granulomas since these are more likely to yield an organism. (1,7,9,10) Other features of the tissue reaction often provide a clue to the type of organism present. For example, the presence of neutrophils in granulomas should prompt a search for Blastomyces, while the association of eosinophils with granulomas often indicates the presence of Coccidioides. "Infarctlike" necrosis may be seen in granulomas caused by Histoplasma or M tuberculosis, while a bubbly appearance of the cytoplasm of histiocytes and multinucleated giant cells is a clue to the presence of Cryptococcus. Despite these helpful associations, there is enough histologic overlap in tissue reactions that no one feature is absolutely specific for a particular organism.

Examining Special Stains for Organisms.--The search for organisms must always begin with hematoxylin-eosin (H&E)-stained sections. Many pathologists, incorrectly assuming that organisms will not be visible on H&E, skip this vital step and go directly to the special stains. Although this is true for mycobacteria, most fungi are readily identified on careful examination of an H&E stain. Examination of H&E-stained sections also enables the pathologist to discern subtle differences in morphology between organisms and place the organism in the context of the associated tissue reaction. The notable exception to this general rule is Histoplasma, which is virtually impossible to detect within necrotizing granulomas on H&E-stained sections. (10-13)

The histochemical stains used most often for identification of organisms are GMS for fungi and ZN (often colloquially called "AFB" for acid-fast bacteria) for mycobacteria. Although some laboratories additionally perform a periodic acid-Schiff stain for fungi, the GMS stain is preferable, primarily because periodic acid-Schiff provides less contrast between fungi and background debris. (4,7,14,15) Immunofluorescent and immunohistochemical techniques are more specific but have their limitations, (7,16) the most important of which is lack of widespread availability. For mycobacteria, an alternative to the ZN stain is the auramine/auramine-rhodamine fluorescence technique, which has been shown to be equivalent to culture and superior to conventional acid-fast stains in sensitivity. (1,9,17,18) When H&E-stained sections have been examined, we recommend examining the GMS-stained sections before the more time-consuming ZN-stained sections. Scanning at a relatively low magnification (x10 ocular, x20 objective) suffices to pick out most organisms, although Histoplasma may be missed without a careful search with a x40 objective. If results with the GMS stain are negative, the pathologist should then spend more time examining the ZN-stained sections, hunting with a x40 objective for mycobacteria, which are often few and difficult to find.

The morphologic features of the common fungal causes of lung granulomas are compared in Table 3. It is important to remember that GMS is not a specific stain for microorganisms, as it often stains mucin, elastic tissue, and dust particles in the background lung. Dust and pollen particles pose a particular challenge because they may mimic the appearance of fungal yeasts. However, these particles are smaller in size and lack the characteristic shapes and budding forms associated with fungal yeasts. Mycobacterial organisms may occasionally be seen with GMS stains in certain cases and should not be dismissed as an artifact. It is also important to note that, on occasion, a poorly performed GMS stain may yield falsely negative results; thus, it is important to check standard tissue controls in every case. Histoplasma yeasts in necrotizing granulomas pose a particularly difficult problem because they may stain very weakly with GMS and be overlooked. (7,10) Fungi of any type may also be missed if they are few.

Regarding the ZN stain, the most important point to remember is that, in most cases, mycobacteria are few and difficult to find, partly because of the use of xylene in routine processing. (19) Therefore, cursory scanning at low magnification will miss most my cobacteria. We recommend spending at least a few minutes at high magnification (310 ocular lens, 340 objective) hunting for mycobacteria in the necrotic area of each necrotizing granuloma, constantly adjusting the fine focus to ensure detection of organisms that appear only on certain planes. Other authors go further and use a high-power oil immersion objective. (13,20) Organisms are by far more common in the center of the necrosis (10,13) but may occasionally be found in the periphery of the necrosis or even within the cellular granulomatous rim. It is not uncommon to examine several large necrotizing granulomas and find only a few mycobacteria. When mycobacteria are identified, the next step, for the purpose of choosing appropriate antibiotic therapy, is to differentiate between tuberculous and nontuberculous mycobacteria. Unfortunately, the morphologic appearance of mycobacteria on histologic sections is not reliable for speciation. The published literature on speciation of mycobacteria by using microscopic morphologic features of the organisms is based mostly on smears made from microbiologic cultures rather than formalin-fixed, paraffin-embedded histologic material. Histologic studies that claim distinctive morphologic features for particular mycobacterial species have included only nontuberculous mycobacteria, (21-23) and no data exist to show the accuracy of mycobacterial subtyping by pathologists blinded to culture results. Even in smears made from microbiologic culture material, morphologic features such as cording, beading, or size, although associated with certain species, do not allow accurate or definitive speciation of mycobacteria. (24-26) Currently, the only definitive methods of mycobacterial speciation are microbiologic culture and molecular methods such as the polymerase chain reaction (PCR) (see below). In most cases, speciation is not a problem because culture test results are also positive. In fact, mycobacteria often grow in cultures even when special staining of histologic material shows negative results. (1,10,18,23,27,28) When results with histologic special stains are positive but those of cultures are negative, (10,18,28,29) or when biopsied tissue was not submitted for culture, PCR is the only means of determining the species of the organism. Finally, if PCR yields a negative result or is unavailable, the clinician may elect to initiate empiric therapy.

Role of PCR and Other Molecular Methods for Detection and Speciation of Mycobacteria.--Pathologists and clinicians are sometimes faced with the difficult situation whereby mycobacteria are identified by histologic methods but speciation is not possible because no specimen was submitted for culture or culture results are negative. In an attempt to find a solution to this problem, the role of PCR and other molecular methods in the detection and speciation of mycobacteria in formalin-fixed, paraffin-embedded tissue has come under intense investigation. (19,30,31) In general, these studies show that PCR is as sensitive as microbiologic cultures for the detection of mycobacteria in formalin-fixed tissues and is more sensitive than ZN staining. (19) It is also possible to determine the species of organisms by this method. Currently, however, PCR for detection and speciation of mycobacteria in formalin-fixed, paraffin-embedded sections is not in routine diagnostic use in most laboratories and remains restricted to a few reference centers.

How to Report the Presence of Organisms.--The subtyping of infectious mycobacterial or fungal disease (acute histoplasmosis versus histoplasmoma, progressive primary tuberculosis versus secondary tuberculosis, etc.) often requires clinical and radiographic information that is usually unavailable to the pathologist. Therefore, for the purposes of the pathology report, describing the tissue reaction and stating the organism present is sufficient in most instances. A pathologic diagnosis for a case in which Histoplasma yeasts are identified within a necrotizing granuloma would be described as "necrotizing granuloma (Histoplasma identified)." If more clinical information is available, a more specific diagnosis may be rendered. For example, if Histoplasma yeasts are identified in a necrotizing granuloma and the pathologist is provided with the information that the lesion is an incidentally detected solitary lung nodule, a diagnosis of histoplasmoma can be rendered.

Step 2: Identifying Histologic Features of Noninfectious Granulomatous Lung Diseases

The key diagnostic features of the common noninfectious granulomatous lung diseases are listed in Table 2. These are the major features that are required for a pathologic diagnosis. In the absence of these features, a definitive pathologic diagnosis is usually not possible and a descriptive diagnosis must suffice. Note that some diagnoses require clinical input, while others can be rendered (or suggested) solely on the basis of their histologic features. We recommend using Table 2 to quickly narrow the differential diagnosis. Details of the pathologic findings and the differential diagnosis can then be obtained from the text (see below).

Step 3: Review of Special Stains and Descriptive Diagnoses

What to Do When Organisms Are Not Found in a Granuloma.--Often, organisms are not found within granulomas despite a meticulous search. Even in necrotizing granulomas, this is a fairly frequent scenario. (10,13) In such cases, the most productive next step for the pathologist is to reevaluate the special stains: reexamination of the GMS-stained sections, in particular, results in the detection of several cases of initially overlooked Histoplasma infection (S.M., unpublished data,March 2009). If the necrotic portion of the granuloma is not represented on the slide with the special stain, it may be productive to recut the block and repeat the stain. In a study by Goodwin and Snell, (32) 10 of 17 histoplasmomas were diagnosed only after an effort was made to include the necrotic center in a recut. If some blocks with necrosis are not initially stained with special stains, staining these may also be productive. Ulbright and Katzenstein (10) showed that examining 2 blocks with necrosis is adequate in most cases, with the caveat that examination of more sections is probably indicated when the histologic appearance is especially suggestive of infection or when a specific noninfectious diagnosis is being considered.

Despite these steps, a substantial proportion of necrotizing granulomas remain unexplained. Ulbright and Katzenstein (10) suggested that such cases might represent infectious granulomas in which the organism has been killed or removed by the inflammatory reaction. In such cases, we recommend issuing a descriptive diagnosis including the presence/absence of necrosis and the absence of identifiable organisms. In the case of necrotizing granulomas, acomment such as "the etiology is most likely infectious despite negative special stains" may be appropriate. In some cases in which organisms cannot be found by histologic examination, results of cultures may be productive. From surgically excised granulomas, mycobacteria (mainly nontuberculous) often grow in culture despite negative ZN and auraminerhodamine staining results on histologic material.1 Finally, clinicians may classify a small number of unexplained necrotizing granulomas as histoplasmosis or anti-neutrophil cytoplasmic antibody (ANCA)-related vasculitides on the basis of serologic studies.

MYCOBACTERIA

Tuberculosis

Worldwide, Mycobacterium tuberculosis outnumbers nontuberculous mycobacteria and fungi such as Histoplasma as the leading cause of granulomatous lung disease. Considering the prevalence of the disease, there are surprisingly few papers in the current literature on the histologic features of culture-proven pulmonary tuberculosis. The few histologic studies available are mainly from the years when the mere presence of necrotizing granulomas was considered diagnostic of tuberculosis. (13,33) In fact, before 1953, all rounded granulomas of the lung were assumed to be tuberculomas. (13,32)

The granulomas of tuberculosis are typically necrotizing (Figure 1, A) but may be nonnecrotizing or a mix of both types. (20,34) As discussed in step 3 above, acid-fast organisms (Figure 1, B) may be few and difficult to find. The granulomas of tuberculosis may be randomly located or bronchiolocentric, although it is important to remember that granulomas may involve bronchi and bronchioles in virtually any infection, (35) as well as sarcoidosis and hypersensitivity pneumonitis. The granulomas of tuberculosis may also involve blood vessels, although less frequently than in sarcoidosis. (36) The features may remotely resemble a true primary vasculitic disorder, such as Wegener granulomatosis. The histologic appearance of tuberculous granulomas may be indistinguishable from those of nontuberculous mycobacterial infection. (28,37,38) This was most elegantly demonstrated in a 1963 study by Corpe and Stergus (37) in which 27 pathologists with expertise in mycobacterial disease were blinded to culture results and asked whether histologic changes on 25 slides were due to M tuberculosis or Runyon group III nontuberculous mycobacteria. In most cases, the pathologists could not differentiate between the 2 infections, and when they felt they could, they were often mistaken. Fungi such as Histoplasma and Coccidioides may also produce a tissue response identical to tuberculosis. (7) Because the histologic features of tuberculosis are not organism-specific, the diagnosis rests on detection (and subsequent speciation) of mycobacteria. Issues relating to identification and speciation of mycobacteria are discussed in detail in "Examining Special Stains for Organisms" (see 2nd page of this article).

[FIGURE 1 OMITTED]

Nontuberculous Mycobacteria

In the United States, NTM are increasingly being recognized as an important cause of lung disease. In the past, lung infection by NTM was thought to occur mainly in the setting of immunodeficiency or preexisting lung disease, such as chronic obstructive pulmonary disease or cystic fibrosis. It is now well known that NTM-related lung disease also occurs in immunocompetent individuals without preexisting lung disease. (39)

In immunocompromised patients, such as those with the acquired immunodeficiency syndrome (AIDS), NTM infection is characterized by collections of mycobacterialaden foamy histiocytes, poorly formed granulomas, or the lack of any significant inflammatory response. (40) Mycobacteria in this form of NTM disease are numerous and easy to find, and culture results are usually positive.

In immunocompetent patients, NTM infection has been associated with a wide variety of histologic findings, including granulomatous inflammation indistinguishable from tuberculosis (Figure 1, C and D, and "Tuberculosis" above). (28,37,38) Like tuberculosis, both necrotizing and nonnecrotizing granulomas may be found (23,28) and the granulomas may be peribronchiolar. Cases with nongranulomatous inflammation have also been described, but since NTM are frequent colonizers of the respiratory tract, it is debatable whether NTM infection in such cases is etiologic or merely coincidental. (23,28) Nontuberculous mycobacteria infection may also be associated with airway disease, manifested histologically as bronchiectasis or chronic bronchiolitis. (41,42) Nontuberculous mycobacteria have been isolated from cases of the middle lobe syndrome, suggesting a predilection for this location. (43) Definitive diagnosis of NTM disease rests on identification and speciation of mycobacteria (see "Examining Special Stains for Organisms" on the 2nd page of this article).

Finally, exposure to aerosolized NTM can cause a hypersensitivity pneumonitis-like illness known as "hot tub lung." In keeping with current concepts of the pathogenesis of this condition, this entity will be discussed in the section on noninfectious granulomatous lung disease (below).

FUNGI

The tissue reaction to the common pulmonary fungi discussed in this section varies greatly, depending on the size of the inoculum, the duration since exposure, and the immunologic status of the host. The sequence of events and the wide variety of inflammatory responses and outcomes are similar in many respects to mycobacterial infections. In most immunocompetent individuals, exposure to a small inoculum leads to asymptomatic selflimited infection. Heavier exposure may lead to an acute flulike or pneumonia-like illness (acute pulmonary histoplasmosis, acute blastomycosis, or acute coccidioidomycosis/ valley fever). (44-50) Such cases may be diagnosed clinically as influenza or community-acquired pneumonia. When the correct diagnosis is made, it is usually based on serologic rather than histologic findings. (49,51) Since a biopsy is rarely performed in acute disease, the pathologic findings remain poorly described. Both asymptomatic infection and acute symptomatic disease clear completely in most cases.

In some individuals, healing is not accompanied by complete clearance of organisms. Instead, organisms persist in a walled-off nodule characterized by a well-formed, necrotizing granuloma similar to a tuberculoma (eg, histoplasmoma, cryptococcoma, coccidioidoma). (12-14,32,45,52,53) This is the most common form of granulomatous fungal disease encountered by surgical pathologists. These lesions are biopsied or resected because they form nodules, which are difficult to separate from malignant tumors on clinical and radiographic grounds.

In some patients, usually those with an underlying predisposing illness, infection progresses instead of being cleared or contained. This results in chronic symptomatic fungal lung disease (eg, chronic pulmonary histoplasmosis, chronic blastomycosis, persistent coccidioidal pneumonia). (45,54) The pathologic manifestaton of this form of fungal lung disease is complex and includes necrotizing granulomas in combination with evidence of the underlying predisposing illness (such as emphysema) or superimposed complications (such as cavities).

The final piece in the spectrum of fungal lung disease does not typically involve granulomatous inflammation but is mentioned here to give the reader a sense of the gamut of tissue responses to fungal organisms. This type of disease occurs in immunocompromised individuals, who often develop disseminated infection characterized by unchecked proliferation of organisms accompanied by little or no inflammatory response (eg, disseminated histoplasmosis, disseminated blastomycosis, disseminated cryptococcosis, disseminated coccidioidomycosis). (55,56) Granulomas are absent, and if present, are poorly formed. This form of disease is often widespread, with a predilection for the lymphohematopoietic system, but it may also involve the lungs.

Histoplasma

Lung involvement in Histoplasma infection takes many forms, but the tissue reaction can be broadly divided into 3 main types. The first is an intra-alveolar lymphohistiocytic infiltrate with small granulomas and variable necrosis, seen in acute pulmonary histoplasmosis. The second is well-formed necrotizing granulomatous inflammation, exemplified by histoplasmomas but also seen in chronic pulmonary histoplasmosis. (12-14,32,54) The third type of tissue reaction consists of sheets of histiocytes within the interstitium, packed with numerous organisms, as seen in disseminated histoplasmosis. (55) By far the most common form of Histoplasma-related lung disease encountered by surgical pathologists is a histoplasmoma (Figures 1, E and F, and 2, A and B). The classic lesion is a large nodule with abundant central necrosis surrounded by a thin rim of epithelioid histiocytes and a fibrotic capsule of variable thickness (Figure 1, E). Small calcific particles may be present within the necrosis or the calcification may be more irregular. Overall, the histologic appearance is identical to a tuberculoma or coccidioidoma. (13) Blood vessels in and around the granuloma may show a prominent nonnecrotizing vasculitis, as noted previously for mycobacterial disease. (10,54) It is important to note that the "vasculitis" in histoplasmosis and other granulomatous infections is a secondary phenomenon related to the granulomatous inflammation, in contrast to a true primary vasculitis such as Wegener granulomatosis. In describing cases with such vascular changes, it is advisable for pathologists to avoid the term vasculitis in their reports because clinicians may misconstrue this as a true primary vasculitis. The vascular changes in Histoplasma infection often result in parenchymal necrosis with an infarctlike quality; that is, ghosts of alveoli can be discerned within the necrotic areas. (14,32,54,57) As with other infections, smaller satellite, nonnecrotizing granulomas and areas of organizing pneumonia may be scattered in the lung parenchyma around the main necrotizing granuloma. With time, the necrotic center is progressively replaced by fibrosis and calcification.

[FIGURE 2 OMITTED]

One key point to remember is that the visibility of Histoplasma on H&E-stained sections depends on the histologic context. While Histoplasma is readily identifiable within macrophages in disseminated histoplasmosis, it is not visible on H&E-stained sections within necrotizing granulomas (histoplasmomas).10-13 Although organisms are present within the necrotic areas, they cannot be resolved from the background necrotic debris on H&E-stained sections (Figure 2, A). Structures that at first glance appear to be Histoplasma yeasts are invariably small microcalcifications, which are common in old necrotizing granulomas of any etiology. (10,11) They can be distinguished from fungal yeasts because they are usually GMS negative, basophilic, and pleomorphic. Demonstration of Histoplasma in necrotizing granulomas is best accomplished by performing GMS staining. The presence of yeast forms that are not appreciated on H&E, but visible on GMS, is virtually pathognomonic of Histoplasma. With a GMS stain, Histoplasma yeasts are small, uniform, and oval (Figures 1, F, and 2, B). A characteristic feature is that some taper to a point at one end or sometimes both ends. They are scattered within the necrotic center of the granuloma and may be present in clusters or singly. Narrow-based budding, although often cited as a characteristic finding, is not always identifiable, especially when organisms are few. Histoplasma yeasts tend to be a mixture of lightly stained, hollow-appearing forms with thin, delicate outlines and more darkly stained solid forms, with the staining characteristics varying with the strength of the stain in an individual case. (10,11) Rare hyphal forms have also been described. (11,58) Histoplasma seldom grows in culture from histoplasmomas probably because the organisms are nonviable in most of these lesions. (10,13,59) Therefore, histopathologic examination is often the only means of confirming the diagnosis. (14)

Histoplasma may be confused with Cryptococcus, which can be associated with an identical granulomatous response. Small, capsule-deficient forms of Cryptococcus pose a particular challenge. A helpful morphologic feature in the differential diagnosis is that Histoplasma cannot be seen (in necrotizing granulomas) on H&E-stained sections, whereas Cryptococcus yeasts can (Figure 2, A and C). Both organisms are highlighted by the GMS stain, whereby the main differentiating features are the uniform size, oval shape, and occasional tapered forms of Histoplasma as opposed to the nonuniform size, round shape, and lack of tapering of Cryptococcus (Figure 2, B and D).

Histoplasma may also be confused with Pneumocystis. Fortunately, this is not a common problem because, in most cases, Pneumocystis pneumonia is characterized by an intra-alveolar frothy exudate, while Histoplasma infection is characterized by a necrotizing granuloma. However, in 5% to 17% of cases, Pneumocystis may elicit a granulomatous response (60-62) (Figure 2, E). Such cases, termed granulomatous Pneumocystis pneumonia, can mimic acute pulmonary histoplasmosis, since both conditions feature granulomas. While the frothy intraalveolar exudate of Pneumocystis differs from the fibrinous intra-alveolar exudate of acute pulmonary histoplasmosis, it is not always present. Morphologic appearance of the organisms with GMS is the key differentiating feature. Although Histoplasma and Pneumocystis are similar in size, Pneumocystis cysts are round rather than oval, are often crescent-shaped or sickle-shaped (Figure 2, F), and do not bud or taper. In difficult cases, results of microbiologic cultures or immunohistochemical stains may be helpful.

Cryptococcus

Cryptococcosis of the lung includes a wide spectrum of tissue reactions (63-65) that depend on the immunologic status of the host. Immunocompetent patients tend to develop granulomas, whereas the reaction is more variable in immunocompromised patients. The typical granulomatous reaction to Cryptococcus is confluent nonnecrotizing granulomatous inflammation with numerous multinucleated giant cells and scattered chronic inflammation (Figures 2, C, and 3, A). Multinucleated giant cells may predominate over granulomas. At low magnification, the presence of engulfed Cryptococcus yeasts imparts a bubbly appearance to the cytoplasm of histiocytes and multinucleated giant cells. At high magnification, Cryptococcus yeasts can be identified by careful examination within multinucleated giant cells and granulomas (Figures 2, C, and 3, A). Cryptococcus may also be found within necrotizing granulomas (cryptococcomas) similar to those seen with mycobacterial or other fungal infections. In this setting, too, organisms are identifiable, both within the necrotic areas and the surrounding granulomatous rim. (10) Finally, in immunocompromised patients, innumerable Cryptococcus yeasts grow in sheets within alveolar spaces, alveolar septa (interstitium), and alveolar septal capillaries and bronchioles, with little or no inflammatory reaction. (65-67) Not surprisingly, patients with AIDS occasionally have concurrent Pneumocystis infection. (66)

On H&E-stained sections, Cryptococcus yeasts are round with well-defined but pale-staining blue-gray walls (Figures 2, C, and 3, A). Because of the pale-staining walls, the organisms may easily be overlooked at low magnification. (7) Characteristically, the organism retracts from the cytoplasm of the cell that engulfs it, resulting in the formation of a halo around the organism. Grocott methenamine silver stains the organisms well and, as with other fungi, often highlights far more organisms than initially appreciated (Figures 2, D, and 3, B). The capsule of Cryptococcus stains deep red with mucicarmine and this feature may be used to support the diagnosis. (7,66) However, it is important to remember that absence of mucicarmine staining does not exclude Cryptococcus because the organism may lack a capsule ("capsule-deficient Cryptococcus"). (66,68-70) In such cases, the Fontana-Masson stain is useful because it stains the cell wall of Cryptococcus, including that of capsule-deficient forms. (15,70)

The differential diagnosis includes Histoplasma and Blastomyces. (70) For details of the features useful in the differential diagnosis, see Table 3 and the sections on Histoplasma and Blastomyces.

Coccidioides

The most common form of Coccidioides infection encountered by the surgical pathologist is the necrotizing granuloma (coccidioidoma). (45,52) Eosinophils may be numerous, (71) scant, or absent and neutrophils may be prominent. As with other infections, the granulomas may be peribronchiolar or may communicate with or destroy bronchioles. An accompanying nonnecrotizing vasculitis, as noted previously with other infections, may be present. (10) Smaller satellite, nonnecrotizing granulomas are often seen in surgically resected specimens. The overall picture is indistinguishable from granulomas caused by mycobacteria or other fungi. (13) Demonstration of organisms is therefore essential for diagnosis.

[FIGURE 3 OMITTED]

Coccidioides organisms are most often found within the necrotic centers of necrotizing granulomas, although they show less of a propensity for the center than Histoplasma and may also be found within nonnecrotizing granulomas. The organisms consist of large, thick-walled, spherical structures ("spherules") filled with smaller yeastlike structures ("endospores") (Figure 3, C and D). Not uncommonly, spherules are found in a ruptured, fragmented, emptying or empty state. Endospores of various sizes may lie scattered in the necrotic debris, mimicking other fungal yeasts. Neither spherules nor endospores show budding. Although some authors have reported difficulty in detecting organisms on H&E stained sections, (45) they are usually easily found. (7,52) As with other fungi, the GMS stain reveals more organisms than initially appreciated on H&E-stained sections. Grocott methenamine silver highlights the spherules as well as the endospores. A mycelial form of the fungus, consisting of septate hyphae with arthrospores, may also be encountered if there is communication with ambient temperature. (45,52) As with Histoplasma, the organism may not grow in cultures (especially from coccidioidomas) (10); in such cases, histologic examination may be the only means of establishing the diagnosis.

If spherules and endospores are both present (Figure 3, C and D), the diagnosis is straightforward. However, in many cases it may be difficult to find these characteristic fungal structures. In the absence of one or the other form, differentiation from Blastomyces can be difficult (see Table 3 and "Blastomyces" below). Size of the organism is a helpful feature because Coccidioides spherules are usually larger than Blastomyces yeasts. Broad-based budding is a feature of Blastomyces but not of Coccidioides. In difficult cases, correlation with the results of microbiologic cultures is prudent.

Blastomyces

Infection with Blastomyces is uncommon. It must be suspected when granulomas or multinucleated giant cells are accompanied by acute inflammation (1,6-8,72) (Figure 3, E and F). Classic cases show granulomas with frankly suppurative centers, in contrast to the pink or slightly "dirty" necrosis seen in granulomas due to other organisms. As with other infections, the granulomas may be bronchiolocentric. (35) The large, thick-walled yeast forms of Blastomyces can be identified on H&E-stained sections, although they may be few and difficult to find. Broad-based budding is fairly characteristic but not always present. On H&E-stained sections, nuclear material (multiple nuclei) can often be identified within the yeasts; this can mimic the endospore-filled spherules of Coccidioides (compare Figure 3, C, and 3, E).

Blastomyces yeasts are larger than the yeasts of Histoplasma and Cryptococcus and the endospores of Coccidioides but smaller than Coccidioides spherules (Table 3); however, enough overlap and size variation exists to cause potential confusion between these organisms (7,73,74) (Figure 3, A through F). Mucicarmine staining, while characteristic of Cryptococcus, can also be seen with Blastomyces, although it tends to be weaker in the latter (8,74) and has been attributed by some authors (7) to overstaining. Multinucleation and broad-based budding are features of Blastomyces that are absent in Cryptococcus. Blastomyces differs from Histoplasma in that it is visible on H&E-stained sections, has a thick wall, broad-based budding, multiple nuclei, and a suppurative tissue reaction. Differentiation of Blastomyces from Coccidioides can be difficult. For features helpful in separating the two, see Figure 3, C through F; Table 3; and "Coccidioides" above.

Pneumocystis

Most pathologists are familiar with the usual picture of Pneumocystis pneumonia, which is a frothy, eosinophilic, intra-alveolar exudate accompanied by mild interstitial chronic inflammation. In 5% to 17% of cases (61,62) a granulomatous reaction is encountered (Figure 2, E). In most, it consists of poorly formed intra-alveolar granulomas characterized by epithelioid histiocytes palisading loosely around an eosinophilic exudate. Some cases may show intra-alveolar nonnecrotizing granulomas without an exudate, (75) scattered multinucleated giant cells, organizing granulomatous pneumonia, (60) or well-formed granulomas with or without central necrosis.

In immunocompromised patients, the above histologic appearance is highly suggestive of granulomatous Pneumocystis pneumonia and calls for especially close scrutiny of the GMS-stained section. A frothy intra-alveolar exudate should also raise the possibility of Pneumocystis, although it is not always present. Organisms can be scarce and difficult to find. They consist of small, round, sickle-shaped or crescent-shaped cysts (Figure 2, F). Differentiation from acute pulmonary histoplasmosis is discussed in the section on Histoplasma.

Aspergillus

Aspergillus may cause invasive, saprophytic, or allergic lung disease, (76) depending primarily on the immune status of the host. The 2 most common forms of lung involvement by Aspergillus (aspergilloma and invasive aspergillosis) do not feature granulomas. However, granulomas are a prominent feature of 2 less common forms of pulmonary aspergillosis (chronic necrotizing pulmonary aspergillosis and allergic bronchopulmonary aspergillosis).

In the rare condition known as chronic necrotizing pulmonary aspergillosis, Aspergillus infection results in semi-invasive, chronic, indolent, cavitary disease in patients who have preexisting chronic lung disease or are mildly immunocompromised. (77) The histologic picture is characterized by necrotizing granulomas containing Aspergillus hyphae. (78) The presence of granulomas is indicative of limited tissue invasion. The granulomas may cause extensive parenchymal consolidation, may lead to bronchiectatic cavities, or may be exclusively bronchocentric. The bronchocentric cases may be accompanied by nonnecrotizing granulomas in a lymphangitic distribution. A nonnecrotizing vasculitis can occur. Eosinophils are not prominent.

Chronic necrotizing pulmonary aspergillosis should be differentiated from other forms of Aspergillus-related lung disease. In mycetomas (aspergillomas), the fungal organisms do not invade the surrounding parenchyma or evoke a granulomatous response. Allergic bronchopulmonary fungal disease (see below) occurs in persons with asthma and is easily differentiated from chronic necrotizing pulmonary aspergillosis by the prominence of tissue eosinophils (see below). Invasive aspergillosis differs from chronic necrotizing pulmonary aspergillosis because it features vascular invasion by fungal hyphae and extensive parenchymal necrosis without granuloma formation.

Allergic bronchopulmonary aspergillosis is a noninvasive form of Aspergillus lung disease characterized by a hypersensitivity response to Aspergillus antigens, occurring mostly in asthmatic patients. (79-81) Less commonly, other fungi such as Curvularia and Candida may cause similar morphologic changes. (82) Hypersensitivity to the fungus results in a distinctive tissue reaction characterized by the triad of mucoid impaction of bronchi, bronchocentric granulomatosis, and eosinophilic pneumonia (Figure 4, A through D). Mucoid impaction of bronchi is characterized by obstruction of proximal bronchi by large, laminated, gelatinous mucus plugs filled with viable and necrotic eosinophils, neutrophils, fibrin, and necrotic debris (Figure 4, A). Charcot-Leyden crystals are often present, reflecting the eosinophil-rich infiltrate (Figure 4, B). Aspergillus hyphae may be found within the mucus (Figure 4, B [inset]). Bronchocentric granulomatosis is characterized by necrotizing granulomas centered exclusively on bronchi and bronchioles distal to the bronchi affected by mucoid impaction (Figure 4, C). The necrotic centers of the granulomas are rich in eosinophils and necrotic debris (Figure 4, D). Eosinophilic pneumonia is usually a focal finding in allergic bronchopulmonary aspergillosis. The main histologic feature is filling of the alveolar spaces with eosinophils. Histiocytes often accompany the eosinophil-rich infiltrate. The interstitium is usually thickened by a combination of eosinophils, lymphocytes, and plasma cells. A mild nonnecrotizing vasculitis may be present. The pathologic diagnosis of allergic bronchopulmonary aspergillosis is confirmed by the identification of Aspergillus hyphae scattered within the lumens of bronchi or bronchioles (Figure 4, B [inset]). Organisms may be sparse, fragmented, and difficult to demonstrate, even with a GMS stain. Even in the absence of demonstrable fungi, the triad of histologic features described above strongly suggests the diagnosis. The differential diagnosis is broader when individual components of the triad are present in isolation.

Several entities enter the differential diagnosis, including (1) other forms of Aspergillus lung disease; (2) mucoid impaction of bronchi, bronchocentric granulomatosis, or eosinophilic pneumonia, occurring in isolation; and (3) diseases due to other fungi with septate hyphae. Differentiation from other forms of Aspergillus lung disease is discussed in the preceding section on chronic necrotizing pulmonary aspergillosis. Each of the features of the histologic triad of allergic bronchopulmonary aspergillosis can occur in isolation in other settings. Mucoid impaction of bronchi can occur in cystic fibrosis or chronic bronchitis. (83) Bronchocentric granulomatosis may occur in other infections. (35) Eosinophilic pneumonia occurs in several other settings, including drug reactions, parasitic and coccidioidal infection, Churg-Strauss syndrome, and idiopathic disease. Therefore, it is a combination of features, rather than any individual finding, that raises the possibility of allergic bronchopulmonary aspergillosis. Definitive diagnosis requires identification of fungal hyphae. The classic appearance of Aspergillus (septate hyphae with narrow-angle branching) is well known to most pathologists, but it must be stressed that this appearance is not specific for Aspergillus. (7) It can occasionally be seen in other fungi such as Pseudallescheria (84,85) and Fusarium. Cultures are required for definitive identification.

PARASITES

Dirofilaria

Dirofilaria (the dog heartworm) is the most common parasite associated with a granulomatous reaction in the lung. The organism is a filarial nematode that typically infects dogs but may rarely infect humans via a mosquito vector. Human pulmonary dirofilariasis is rare. Almost all reported cases in the United States have been from the eastern half of the country and Texas. (86,87) The worm resides in the right side of the heart, from where it embolizes to the lung; there, it is most often found within a thrombosed pulmonary artery surrounded by infarct-like necrosis. (88) A granulomatous reaction is seen in the adjacent tissue in about one-third of cases, eosinophils are present in around two-thirds, and a nonnecrotizing vasculitis is seen in approximately half. (89) The diagnosis rests on identification of the organism, which is often fragmented, degenerated, and calcified. It is usually large, eosinophilic and oval, with a complex internal structure and a lighter-staining, multilayered, smooth cuticle. (90)

NONINFECTIOUS GRANULOMATOUS LUNG DISEASE

Sarcoidosis

In the United States, sarcoidosis is the most common noninfectious cause of lung granulomas encountered by surgical pathologists. (1) However, it is important to emphasize at the outset that nonnecrotizing granulomas occur in the lung in several conditions other than sarcoidosis and that none of the histologic features listed below are absolutely pathognomonic for sarcoidosis. (91) Histologically, sarcoidosis is characterized by discrete, well-formed, interstitial nonnecrotizing granulomas (Figure 5, A and B). Although nonnecrotizing granulomas predominate, small foci of pink necrosis are occasionally encountered. (1,92,93) An underappreciated feature of sarcoidosis is the lymphangitic ("following the lymphatics") distribution of the granulomas (Figure 5, A). In the lung, lymphatics run in the pleura, interlobular septa, and bronchovascular bundles (bronchi/bronchioles and arteries); it is in these locations that the granulomas of sarcoidosis are found. (93,94) This distribution also explains why transbronchial biopsies are so effective in detecting sarcoid granulomas. (95) Needless to say, a lymphangitic distribution can only be appreciated in a surgical biopsy (or larger specimen), not in a transbronchial biopsy. A frequent finding in sarcoidosis is the presence of intracytoplasmic inclusions (Figure 6, A and B), which are thought to represent (endogenous) products of macrophage metabolism. These include pink spiderlike structures (asteroid bodies) (Figure 6, A), basophilic concentric calcifications (Schaumann bodies), and calcium oxalate crystals. (96) Calcium oxalate crystals and Schaumann bodies are birefringent (Figure 6, B) and must not be mistaken for foreign (exogenous) material. (97) Although these inclusions are more frequent and numerous in sarcoidosis, they are also found in other granulomatous diseases such as hypersensitivity pneumonitis (Figure 6, C and D), chronic beryllium disease, tuberculosis, histoplasmosis, (96) nontuberculous mycobacterial infection, (97) and talc granulomatosis (Figure 6, E and F). In an individual case, therefore, they cannot be used as a specific indicator of sarcoidosis.

[FIGURE 4 OMITTED]

Other frequent findings in sarcoidosis are concentric, lamellated fibrosis around granulomas (Figure 5, B), hyalinization of and within granulomas, granulomatous vasculitis,36,98 absence of granulomas within air spaces, absence of organizing pneumonia, and absence of interstitial inflammation away from the granulomas. (2,3,94)

The differential diagnosis includes infection, hypersensitivity pneumonitis, hot tub lung, and chronic beryllium disease (Figure 5, A through F). As discussed previously, infectious granulomas may be nonnecrotizing, well formed (sarcoidlike), and peribronchiolar. A careful search for organisms is therefore mandatory before suggesting a diagnosis of sarcoidosis. Organisms especially likely to cause a similar histologic picture include mycobacteria and Cryptococcus. Although the presence of necrosis does not exclude sarcoidosis, it should prompt a more vigorous search for organisms. (91) Other features common in infection but absent in sarcoidosis are organizing pneumonia and extensive necrosis or suppuration. Hypersensitivity pneumonitis enters the differential diagnosis because of the presence of nonnecrotizing granulomas and frequent intracytoplasmic inclusions (Figure 6, C and D). It is usually easily differentiated from sarcoidosis because the picture is dominated by interstitial chronic inflammation rather than granulomas (Figure 5, C). Interstitial inflammation also involves alveolar septa away from the granulomas in hypersensitivity pneumonitis, while it is limited to the granulomatous foci in sarcoidosis. A lymphangitic distribution, well-formed granulomas, and granulomatous vasculitis occur in sarcoidosis but not in hypersensitivity pneumonitis. Hot tub lung can be a more challenging problem because granulomas are more prominent than in hypersensitivity pneumonitis and thus closer in appearance to sarcoidosis (Figure 5, E and F). The main discriminant histologic feature is that granulomas of hot tub lung occur predominantly within air spaces (mainly the lumens of small bronchioles), while those of sarcoidosis are intersti tial. Lymphangitic granulomas and granulomatous vasculitis favor sarcoidosis, while identification of mycobacteria and a history of hot tub use support a diagnosis of hot tub lung.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

The main role of the pathologist in the diagnosis of sarcoidosis is to exclude other etiologies. First, the specimen should be carefully examined for organisms. Second, an attempt should be made to identify features that are not consistent with sarcoidosis. (99) Such features (used to argue for an alternative diagnosis) include granulomas within alveolar or bronchiolar air spaces, organizing pneumonia, interstitial inflammation in alveolar septa away from the granulomas, extensive necrosis or suppuration, numerous eosinophils, or vegetable material. If these features are present, sarcoidosis is unlikely, and this should be communicated clearly to the clinician.

In a surgical biopsy (or larger specimen) with well-formed nonnecrotizing granulomas distributed along lymphatic pathways, in the absence of features that would argue against sarcoidosis, it is reasonable for a pathologist to state that the features are "consistent with sarcoidosis." In most transbronchial biopsy specimens, however, many of the above features cannot be evaluated, and therefore more caution must be exercised. It is acceptable with these specimens to simply diagnose "nonnecrotizing granulomas" and indicate in a comment that no organisms are identified. One could also outline the differential diagnosis or comment that the features are consistent with sarcoidosis in the appropriate clinical context, while realizing that the final diagnosis of sarcoidosis lies with the clinician, who has access to the results of cultures and knowledge of the clinical and radiologic settings. Even armed with this information, the diagnosis of sarcoidosis can be difficult because of the absence of a uniformly accepted clinical or morphologic gold standard.

Chronic Beryllium Disease

Chronic beryllium disease is characterized by a granulomatous reaction in the lung to inhaled beryllium. The histologic picture is characterized by nonnecrotizing granulomas. The appearance is thought to mimic sarcoidosis in almost every respect, including lymphangitic distribution and hilar lymph node involvement. (100) However, illustrations of chronic beryllium disease in the literature show interstitial inflammation and poorly formed granulomas, the overall picture being more similar to hypersensitivity pneumonitis than sarcoidosis. 101,102 In a large histologic series, most cases showed chronic interstitial inflammation with poorly formed to well-formed granulomas. (101) A smaller fraction of cases were described as having well-formed granulomas resembling sarcoidosis, but illustrations of these "sarcoidlike" cases show more interstitial inflammation than usually encountered in sarcoidosis. Like sarcoidosis (and hypersensitivity pneumonitis), inclusions of various kinds (Schaumann bodies, asteroid bodies, crystals) have been documented.

Pulmonary granulomas are part of the diagnostic criteria for chronic beryllium disease. (100) However, to obtain a definitive diagnosis, since the histologic changes are not specific, the clinician should consider the possibility of beryllium exposure, obtain an occupational history, and document sensitization to beryllium by performing a beryllium lymphocyte proliferation test on blood or bronchoalveolar lavage fluid. (103) Exposure to beryllium occurs in occupations that generate beryllium dust, such as primary production of beryllium, metal machining, and reclaiming of scrap alloys; cases have also been documented for individuals living near beryllium facilities. (104) Since the diagnosis needs clinical and laboratory input, the main role of the pathologist is identification of nonnecrotizing granulomas.

Hypersensitivity Pneumonitis

Hypersensitivity pneumonitis is a type IV hypersensitivity reaction of the lung to inhaled organic antigens such as proteins in bird feathers or thermophilic bacteria in moldy hay. The term "hypersensitivity" (or the word allergic in the British equivalent "extrinsic allergic alveolitis") often conjures up images of eosinophils, but in fact eosinophils are scant or absent (105); rather, lymphocytes are the predominant inflammatory cells in hypersensitivity pneumonitis. Plasma cells are also often present. The pathologic findings of hypersensitivity pneumonitis are interstitial (alveolar septal) chronic inflammation with peribronchiolar accentuation, poorly formed interstitial granulomas, and foci of organizing pneumonia. (106,107) This histologic triad is sufficiently characteristic that the pathologist can suggest the diagnosis solely on the basis of microscopic examination. Interstitial chronic inflammation is the most consistent finding and usually dominates the histologic picture (Figure 5, C). It is often accentuated around bronchioles and may be associated with acute or chronic bronchiolitis. Small, poorly formed granulomas or multinucleated giant cells are randomly scattered within the interstitial inflammation and/or bronchiolar walls. The granulomas can be small and difficult to identify and often consist of only a few loosely aggregated histiocytes (108) (Figure 5, D). In some cases, there are only rare multinucleated giant cells in the interstitium without frank granulomas (Figure 6, C and D). Cytoplasmic inclusions (such as asteroid bodies, Schaumann bodies, or cholesterol clefts) are often prominent within granulomas or multinucleated giant cells (Figure 6, C and D). They represent products of macrophage metabolism (endogenous) and must not be confused with foreign material (exogenous) (Figure 6, G and H). The occasional birefringence of the endogenous inclusions seen in hypersensitivity pneumonitis can also lead to misinterpretation as foreign material. Foamy macrophages often accumulate within alveolar spaces in hypersensitivity pneumonitis and are a manifestation of bronchiolar obstruction.

The presence of nonnecrotizing granulomas can raise the possibility of sarcoidosis. In contrast to hypersensitivity pneumonitis, sarcoid granulomas are well formed and have a lymphangitic distribution. Moreover, the lung away from the granulomas in sarcoidosis is normal, while in hypersensitivity pneumonitis there is significant interstitial inflammation even in areas where no granulomas are present (Figure 5, A through D). Granulomatous vasculitis may be present in sarcoidosis but not in hypersensitivity pneumonitis. Finally, small foci of organizing pneumonia are common in hypersensitivity pneumonitis but not in sarcoidosis.

Differentiation from hot tub lung (a hypersensitivity pneumonitis-like disease; see below) is more difficult and may be impossible on histologic grounds. Many of the morphologic features of hypersensitivity pneumonitis and hot tub lung overlap, including mixed interstitial and air space involvement, interstitial chronic inflammation, and organizing pneumonia. In classic cases, however, hot tub lung is characterized by fairly large, well-formed granulomas located predominantly within bronchiolar lumens, as opposed to the small, poorly formed interstitial granulomas of hypersensitivity pneumonitis (Figure 5, C through F). In difficult cases, a final determination of the diagnosis may require a clinical history of hot tub use and identification of mycobacterial organisms (hot tub lung) or a history of other antigenic exposure such as birds or a moldy home (for hypersensitivity pneumonitis). Both conditions may resolve with cessation of exposure to the causative agent.

Lymphoid interstitial pneumonia can be difficult to differentiate from hypersensitivity pneumonitis because it features an interstitial lymphoid infiltrate with loosely formed granulomas. The findings that differentiate lymphoid interstitial pneumonia from hypersensitivity pneumonitis include lack of peribronchiolar accentuation and organizing pneumonia and the frequent presence of an underlying condition such as Sjogren syndrome or human immunodeficiency virus (HIV) infection. Finally, hypersensitivity pneumonitis also needs to be differentiated from other, nongranulomatous causes of chronic interstitial inflammation such as nonspecific interstitial pneumonia or usual interstitial pneumonia.

Hot Tub Lung

Hot tub lung is a recently described entity characterized by a hypersensitivity pneumonitis-like response to nontuberculous mycobacteria (specifically, Mycobacterium avium complex, abbreviated as MAC) inhaled in aerosol form from hot tubs. (39,109-124) Hot tubs provide an ideal temperature for the growth of the organisms and a means for their aerosolization. However, since other sources like shower heads and therapy pools potentially provide similar conditions, (125) the more inclusive term MAC hypersensitivity-like disease has been used. (39) M avium complex organisms are cultured from the sputum, lung tissue, and/or hot tubs in most, but not all, cases. Optimal treatment (cessation of hot tub use and/or corticosteroid or anti-mycobacterial therapy) is unclear, but there is expert consensus that patients should completely avoid reexposure to indoor hot tubs. (39) The condition resolves in most patients regardless of the treatment modality.

The most characteristic pathologic finding of hot tub lung is the presence of granulomas within air spaces (usually the lumens of small bronchioles) (Figure 5, E and F). All other findings are variable: the granulomas may also be located randomly within alveolar spaces or bronchiolar walls, are usually nonnecrotizing but necrotizing granulomas are occasionally present, and are better defined than those seen in hypersensitivity pneumonitis but not as well formed as those of sarcoidosis. Organizing pneumonia is often present. Mycobacteria may or may not be identifiable with acid-fast stains. Interstitial inflammation may be present but is variable in extent. Definitive diagnosis requires a history of hot tub use or, more rarely, exposure to aerosolized water contaminated with mycobacterial organisms. (122,125) In summary, hot tub lung can be suspected, but not definitively diagnosed, solely on the basis of histologic findings.

The main differential diagnostic considerations (Figure 5, A through F) are hypersensitivity pneumonitis (due to causes other than MAC), MAC infection (in patients without hot tubs), and sarcoidosis. The differential diagnosis with hypersensitivity pneumonitis due to other causes is discussed in the section on the latter entity. M avium complex infection in patients without hot tubs may cause an organizing granulomatous pneumonia. Acid-fast organisms may be present (or absent) in both entities. In the absence of a history of hot tub use, therefore, it is impossible to differentiate hot tub lung and MAC infection. Sarcoidosis is easier to distinguish from hot tub lung: it is characterized by better-defined, exclusively interstitial granulomas, a lymphangitic distribution, and granulomatous vasculitis, none of which are features of hot tub lung. Features that favor hot tub lung over sarcoidosis include identification of mycobacteria, air space granulomas, interstitial inflammation away from the granulomas, and organizing pneumonia.

Lymphoid Interstitial Pneumonia

Lymphoid interstitial pneumonia (LIP) is an uncommon form of interstitial lung disease, usually diagnosed in the setting of human immunodeficiency virus/AIDS, Sjogren syndrome, or other diseases involving immune dysregulation. 126-128 Lymphoid interstitial pneumonia is not usually considered a granulomatous disease, but it is mentioned here because small granulomas are commonly part of the histologic picture. Histologically, LIP is characterized by dense and diffuse alveolar septal (interstitial) lymphoplasmacytic chronic inflammation and scattered histiocytes. Germinal centers may be prominent. Small, loosely formed, nonnecrotizing granulomas are often present. (4,129) Immunohistochemical stains show that the inflammatory infiltrate contains more T lymphocytes than B lymphocytes. Gene rearrangement studies demonstrate that the B lymphocytes are polyclonal. (128)

The differential diagnosis includes hypersensitivity pneumonitis and low-grade B-cell lymphomas of extranodal marginal zone/mucosa-associated lymphoid tissue type, both of which may be associated with poorly formed granulomas. The main histologic features differentiating LIP from hypersensitivity pneumonitis are the greater density of the alveolar septal infiltrate and the presence of germinal centers in LIP versus peribronchiolar accentuation of interstitial inflammation and foci of organizing pneumonia in hypersensitivity pneumonitis. Lymphoid interstitial pneumonia must also be differentiated from low-grade B-cell lymphomas. The predominance of T cells over B cells, polyclonality of the B cells, and the absence of a lymphangitic distribution are features that favor LIP.

Wegener Granulomatosis

Ideally, Wegener granulomatosis is diagnosed in patients with upper respiratory tract symptoms, multifocal lung involvement, kidney disease, (130) anti-neutrophil cytoplasmic antibodies (ANCA), (131) and necrotizing granulomas with necrotizing vasculitis. However, this is not always the case in practice. Renal involvement may be absent at presentation, (130,132) the clinical features may overlap with infectious diseases, disease may be limited to the lungs, (133-137) and the lung lesions may be solitary. (134) Test findings with ANCA may be negative (130,131,137,138) or falsely positive. (139) Cases sent for biopsy are often those with atypical clinical features and not infrequently have atypical pathologic presentations. From the surgical pathologist's perspective, relevant clinical history and results of ANCA testing are often unavailable at the time of biopsy. (137) The role of pathologic examination is crucial in such cases and often helps the clinician to initiate therapy in a timely fashion. When classic histologic features are present, Wegener granulomatosis can be diagnosed even in an unusual clinical setting. (134)

On the other hand, if histologic findings are not classic, clinical input and ANCA results assume greater importance (138); cases without necrotizing vasculitis, in particular, should not be diagnosed as Wegener granulomatosis on the basis of pathologic findings alone. In the appropriate setting, the diagnosis can occasionally be made with transbronchial biopsies. (140)

The classic histologic picture of Wegener granulomatosis consists of necrotizing granulomatous inflammation accompanied by necrotizing vasculitis (4,135,136) (Figure 7, A, C, and E). The granulomas are suppurative (neutrophilrich) and resemble abscesses at low magnification (Figure 7, A). The necrotic (suppurative) areas are usually irregular in contour, with a deeply basophilic, "dirty" appearance owing to the presence of neutrophils and nuclear debris (Figure 7, C). Palisading histiocytes, acute and chronic inflammation, and granulation tissue surround the suppurative necrosis. Multinucleated giant cells, when present, are distinctive but not pathognomonic for Wegener granulomatosis. They stand out at low magnification because of the presence of multiple, closely packed, hyperchromatic nuclei. (137) In contrast, compact, "sarcoidlike," nonnecrotizing granulomas are exceptional in Wegener granulomatosis. Eosinophils may be absent or present in small numbers; only rarely are they numerous. (4,132,136,138,141) Hilar lymph nodes do not show granulomas.

Necrotizing vasculitis is the single most important feature in the histologic diagnosis of Wegener granulomatosis but has been poorly defined in the literature. Several points must be stressed regarding its identification.

1. Necrotizing vasculitis can sometimes be very difficult to identify because the affected vessels may be completely necrotic. Since neutrophils in Wegener granulomatosis are commonly necrotic and karyorrhectic, the presence of necrotic neutrophils in the vessel wall is an acceptable surrogate for necrosis of the wall itself. Thus, the vessel-destructive vascular infiltrate that defines necrotizing vasculitis may consist of necrotic neutrophils, (4) a mixture of necrotic neutrophils and histiocytes (Figure 7, E), suppurative necrosis, (132) suppurative granulomas, (136) or fibrinoid necrosis. (132,142)

2. Although a vasculitis comprising lymphocytes or amixture of lymphocytes and histiocytes is very common in Wegener granulomatosis, (132,136) this finding by itself does not constitute necrotizing vasculitis. This type of nonnecrotizing vasculitis is common in infectious granulomas (10) (Figure 7, F). Granulomatous vasculitis without necrosis is also common in sarcoidosis. (36,98)

3. Necrotizing vasculitis is usually found within the inflamed, but nonnecrotic area in the lesion, and this is where it should be sought. Vasculitis, necrotizing or otherwise, is uncommon in the normal lung surrounding the inflamed area. (10,136,142) Within the inflamed area, it is important to identify necrotizing vasculitis in viable parenchyma. Such viable areas are located adjacent to, but not within, the necrosis. Completely necrotic vessels within areas of parenchymal necrosis do not constitute acceptable evidence of necrotizing vasculitis, (4) since it is impossible to determine if they are merely included in the necrosis as "innocent bystanders."

4. Necrotizing vasculitis is most readily identified when it is "eccentric," affecting the vessel focally and leaving the rest of the wall uninvolved. (4,134,137,142) Since the affected vessel is not completely necrotic, recognition of necrotizing vasculitis is facilitated. This type of involvement is common in Wegener granulomatosis.

5. Vasculitis in Wegener granulomatosis affects both arteries and veins; occasionally, capillaries may be involved in a process known as necrotizing capillaritis (132,138,143-145) (see below). Necrotizing capillaritis results in intra-alveolar hemorrhage, which may be massive and potentially fatal.

Other histologic features that may be present in Wegener granulomatosis include small microabscesses, small suppurative granulomas, necrosis of collagen, and organizing pneumonia at the periphery of the main lesion.

Several uncommon histologic variants of Wegener granulomatosis have been described. These are characterized by unusual prominence of 1 histologic feature, often at the expense of the other classic features described above. These include the bronchocentric variant, (146) bronchiolitis obliterans-organizing pneumonia-like variant, (137) eosinophilic variant, (141) and the alveolar hemorrhage and capillaritis variant. (134,144,145,147)

The main differential diagnosis is with infections, (134) which can show necrotizing granulomas, "dirty" necrosis, and vasculitis-like changes (Figure 7, B, D, and F). Solitary lesions, in particular, are far more likely to be infections. (10) A careful search for organisms is therefore mandatory before a diagnosis of Wegener granulomatosis is made. The absence of organisms and presence of a necrotizing vasculitis are the key features required for a pathologic diagnosis of Wegener granulomatosis. Histologic clues that favor infectious granulomas include admixed compact, sarcoidlike, nonnecrotizing granulomas (148) and lymph node involvement, both of which are absent in Wegener granulomatosis. The clinical picture may help if there is a history of renal or upper respiratory tract involvement or ANCA positivity (especially cytoplasmic [c]-ANCA). However, overreli ance on the results of these studies can be dangerous, as an expanding number of infections and other diseases may lead to false-positive results for ANCA tests. (139) As noted above, isolated lung involvement does occur in Wegener granulomatosis, and such cases are more often ANCA-negative than those with severe, active, multi-system disease. In the setting of isolated lung involvement, a case could be made for erring on the side of conservatism rather than overdiagnosing Wegener granulomatosis. (149) A descriptive diagnosis of "necrotizing granulomatous inflammation" is appropriate in such cases. Differentiation of Wegener granulomatosis from Churg-Strauss syndrome is discussed in the following section.

[FIGURE 7 OMITTED]

Churg-Strauss Syndrome

Churg-Strauss syndrome is a rare condition usually diagnosed on the basis of clinical criteria. Histologic support for the diagnosis may come from biopsies of the lung but other sites are more commonly biopsied. The classic histologic picture in the lung consists of necrotizing granulomas, necrotizing vasculitis, and eosinophilic pneumonia, (150) but this characteristic triad is only infrequently present. (151-154) The diagnosis is therefore rarely made on the basis of pathologic findings alone. The granulomas of Churg-Strauss syndrome are well formed, with central necrosis rich in eosinophils. (148) They involve blood vessels as well as bronchioles. The accompanying necrotizing vasculitis is rich in multinucleated giant cells and eosinophils. The surrounding alveolar spaces are filled with eosinophils (eosinophilic pneumonia) and may show evidence of organization. Since the classic triad of findings is rarely encountered on lung biopsy, the pathologist's role in most cases is to recognize and document the presence of tissue eosinophilia and/or necrotizing vasculitis and communicate the significance of these findings effectively to the clinician. (151) A pathologic diagnosis of eosinophilic pneumonia alone (without necrotizing vasculitis) may be sufficient for a clinical diagnosis of Churg-Strauss syndrome in the appropriate setting. Definitive diagnosis requires clinical input and ANCA testing; clinical diagnostic criteria vary widely. (151,154)

The main entities to consider in the histologic differential diagnosis are infection, eosinophilic pneumonia, and Wegener granulomatosis. The combination of prominent eosinophils and vasculitis-like change may be encountered in infection due to parasites such as Dirofilaria and fungi such as Aspergillus or Coccidioides. Allergic bronchopulmonary fungal disease (see above) can mimic Churg-Strauss syndrome because both occur in asthmatic patients and feature eosinophilic pneumonia along with granulomatous inflammation rich in eosinophils. Mucoid impaction of bronchi is a clue to allergic bronchopulmonary fungal disease; GMS-stained sections should be scrutinized carefully for Aspergillus hyphae when this finding is identified. Necrotizing vasculitis is not a feature of allergic bronchopulmonary fungal disease. Eosinophilic pneumonia is far more common in settings other than Churg-Strauss syndrome or as an idiopathic condition. The clinical features of eosinophilic pneumonia in these settings may be difficult, and sometimes impossible, to separate from Churg-Strauss syndrome. (151) In eosinophilic pneumonia, eosinophils fill the air spaces and perivascular inflammation may be prominent. The histologic features that favor Churg-Strauss syndrome over eosinophilic pneumonia (occurring in other settings) include necrotizing granulomas, necrotizing vasculitis, and tissue necrosis without eosinophils. In the absence of these features (especially necrotizing vasculitis), it may be impossible to differentiate Churg-Strauss syndrome from eosinophilic pneumonia on histologic grounds. Clinical features (perinuclear [p]-ANCA positivity and extrapulmonary involvement in Churg-Strauss syndrome) can be helpful in difficult cases. Wegener granulomatosis is usually not difficult to differentiate from Churg-Strauss syndrome. Although the 2 conditions share a combination of necrotizing granulomas and necrotizing vasculitis, neutrophils predominate in the former, while eosinophils are the hallmark of the latter. Although eosinophils may be present in Wegener granulomatosis, only rarely are they present in sufficient numbers to create diagnostic problems. In such cases, the clinical setting is helpful. Unlike Churg-Strauss syndrome, Wegener granulomatosis is frequently associated with destructive upper respiratory tract lesions and c-ANCA positivity but not with asthma or peripheral blood eosinophilia.

Aspiration Pneumonia (Due to Aspiration of Particulate Material)

A variety of substances, including oropharyngeal bacteria, foreign bodies, milk, barium, exogenous lipids, and gastric contents can be aspirated into the lungs, leading to a wide variety of tissue reactions and clinical consequences. (155) Aspiration of oropharyngeal flora leads to the classic acute necrotizing bronchopneumonia known to clinicians as "aspiration pneumonia." Aspiration of gastric acid leads to the acute respiratory distress syndrome, classically occurring in anesthetized patients and characterized histologically by diffuse alveolar damage.

In contrast to these well-characterized conditions, aspiration of particulate gastric contents (food and/or pill fragments) is underrecognized. This entity is particularly relevant to pathologists because it is usually unsuspected clinically (156) and because histologic identification of aspirated material within the lungs is the only means of confirming a definitive pathologic diagnosis of aspiration. The tissue response to the particulate material is quite characteristic. (157-162) The earliest event is acute bronchopneumonia associated with a foreign body giant cell reaction, followed later by organizing pneumonia. In surgical biopsies, organizing pneumonia is often the predominant finding (156,163,164) (Figure 8, A). In the most classic cases, it is associated with acute inflammation and/ or foreign body granulomas or multinucleated giant cells containing aspirated foreign material (156,158,163-170) (Figure 8, B). The presence of particulate foreign material is the key diagnostic finding. It usually consists of vegetable fragments in various stages of degeneration (156,167,171) (Figures 6, G and H, and 8, B). Other forms of aspirated particulate foreign material are less common and include excipients derived from oral pills (talc, microcrystalline cellulose, crospovidone)156 or sodium polystyrene sulfonate. (156,172-174)

The differential diagnosis includes infection, Wegener granulomatosis, and talc granulomatosis. Both infection and Wegener granulomatosis may show a combination of bronchocentric granulomas, acute inflammation, and organizing pneumonia, mimicking the appearance of aspiration. The key diagnostic findings that separate these conditions are necrotizing vasculitis (Wegener granulomatosis), organisms (infection), and vegetable or other aspirated particulate material (aspiration). Aspiration pneumonia due to aspiration of pill fragments containing talc may raise the possibility of talc granulomatosis. The location of the granulomas is important. The granulomas of aspiration pneumonia occur in peribronchiolar parenchyma, while those of talc granulomatosis occur within alveolar septa. Vegetable fragments, organizing pneumonia, and acute inflammation are seen in aspiration but are absent in talc granulomatosis.

[FIGURE 8 OMITTED]

Talc Granulomatosis

The entity of talc granulomatosis was recently in the spotlight due to a controversial case reported in The New Yorker in September 2008. (175) Talc (hydrated magnesium silicate), like microcrystalline cellulose and crospovidone, is an excipient. Excipients are inactive substances used as carriers for the active ingredients of medications. In oral pills/tablets, they provide bulk ("fillers"), hold the ingredients together ("binders"), and cause the pills to dissolve when wet ("disintegrants"). Talc granulomatosis results from intravenous injection of aqueous suspensions of crushed oral medications such as methadone, methylphenidate, tripelennamine, and pentazocine. (176-184) The injected material travels via veins and the right side of the heart to the lungs. There, it is trapped within pulmonary arterioles and capillaries, eliciting a foreign body-type granulomatous response around these vessels (Figure 8, C and D). The perivascular location of the granulomas is not always readily apparent but can be inferred from the alveolar septal distribution. Obstruction of blood flow by the foreign material and the associated granulomatous response may lead to pulmonary hypertensive changes. (182,184) Intravascular thrombi and vascular dilatation may accompany these changes. Talc (and/or microcrystalline cellulose or crospovidone) is identified histologically within the granulomas. Since talc is the most common material found and comprises the bulk of the material in most cases,182 the umbrella term talc granulomatosis, though not entirely accurate, is widely used. Other filler materials, such as microcrystalline cellulose and crospovidone, may also be present. (177,183,185,186) On occasion, the foreign material is scant and the granulomatous response subtle; in such cases, the diagnosis is easily overlooked. A polarizing lens is useful in this situation, since talc and microcrystalline cellulose (but not crospovidone) are strongly birefringent (Figure 8, D [inset]) and stand out at low magnification. Often, far more foreign material becomes evident upon polarization than was evident on H&E. (187)

The presence of interstitial granulomas with birefringent material can raise the possibility of sarcoidosis. The granulomas of sarcoidosis are large, round, and discrete and their distribution is lymphangitic (see "Sarcoidosis") rather than purely alveolar septal. The amount of birefringent material is greater in talc granulomatosis than in sarcoidosis. However, occasional cases of sarcoidosis can show abundant birefringent material,97 and these may be particularly difficult to distinguish from talc granulomatosis. The foreign material in talc granulomatosis is large, sheet/platelike (talc), fiber/rod-like (microcrystalline cellulose), or deep blue and corallike (crospovidone). (177,181,185) In contrast, the inclusions of sarcoidosis are small and crystalline. In difficult cases, acid digestion can be a useful histochemical technique because it removes calcium but not talc. (96) Inhalation (as opposed to intravenous injection) of talc leads to a condition known as inhalation talcosis or talc pneumoconiosis. The pathologic findings of talc pneumoconiosis are poorly described but include interstitial fibrosis and scarring. (188,189) The condition may be difficult to distinguish from talc granulomatosis. As the case reported in the New Yorker illustrates, a history of drug abuse may not be available and a history of occupational exposure may not always settle the issue. Measurement of particle size may be helpful because inhaled talc particles tend to be smaller than injected talc. (190) Aspiration of pill fragments can also be confused with talc granulomatosis. Fortunately, the latter occurrence is quite uncommon (156); it can be differentiated from talc granulomatosis by the peribronchiolar localization of the granulomas and the frequent association with organizing pneumonia and acute inflammation (both absent in talc granulomatosis).

Rheumatoid Nodule

The possibility of a rheumatoid nodule should be considered when a necrotizing granuloma is encountered in the lung of a patient with rheumatoid arthritis. The clinical context is helpful, since most rheumatoid nodules are multiple and subpleural and occur in seropositive patients with active joint disease. (191-193) In the absence of this information, the diagnosis must be made with caution. The histologic appearance is that of a necrotizing granuloma with abundant central necrosis and a rim of palisading epithelioid histiocytes. There is often basophilic karyorrhectic debris at the interface between the necrosis and the granulomatous rim. There may be associated vasculitis (148) but necrotizing vasculitis is not found. The changes described above are histologically indistinguishable from infectious necrotizing granulomas, which, of course, may occur in patients with rheumatoid arthritis; infectious granulomas can also be multiple, bilateral, and subpleural. Therefore, the diagnosis of a rheumatoid nodule is essentially one of exclusion. Test results with special stains for organisms and microbiologic cultures must be negative before a necrotizing granuloma in the lung is presumed to be a rheumatoid nodule.

Bronchocentric Granulomatosis

Bronchocentric granulomatosis refers to the presence of necrotizing granulomas centered exclusively on bronchi and bronchioles. (35,148,194,195) This definition excludes most ordinary granulomatous diseases, which are not limited to bronchi and bronchioles but also involve other components of the lung parenchyma. Some cases that fit this definition are encountered in asthmatic patients as part of the spectrum of allergic bronchopulmonary fungal disease (see Figure 4 and "Aspergillus" above). In such cases, eosinophils are numerous in the necrotic areas. The remaining cases occur in patients without a history of asthma and probably represent undiagnosed infections. 35,196 Tissue eosinophilia is less marked or absent in this group. Rarely, a bronchocentric granulomatosis-type pattern may be found in Wegener granulomatosis. (136,146)

To summarize, bronchocentric granulomatosis is best thought of as an unusual and distinctive tissue reaction seen predominantly in allergic bronchopulmonary fungal disease and various other infections, rather than as the discrete entity that Liebow135 had originally envisioned. When bronchocentric granulomas are associated with prominent eosinophils, a GMS-stained section should be examined carefully and the possibility of allergic broncho-pulmonary fungal disease should be communicated to the clinician. In cases without eosinophils and no history of asthma, the likely infectious nature of the process should be communicated, even if organisms are not detected.

CONCLUSION

In this review, we have outlined a practical, step-by-step approach for surgical pathologists faced with granulomatous lung disease. Several practical tips regarding examination and interpretation of special stains are included to help pathologists detect and identify organisms. The common fungi are illustrated side by side at identical magnification to facilitate recognition of the major differentiating features. Next, we have listed the key features of the noninfectious granulomatous lung diseases in a concise table so that pathologists can rapidly narrow their differential diagnosis and subsequently obtain details of the differential diagnosis from the text. Finally, we address the common situation for which a specific diagnosis cannot be made and suggest practical tips to handle this problem.

Dr Mukhopadhyay would like to thank Anna-Luise Katzenstein, MD, (Syracuse, New York) for allowing him to use slides from her consultation files, from which many of the photomicrographs in this paper are derived.

References

(1.) Woodard BH, Rosenberg SI, Farnham R, Adams DO. Incidence and nature of primary granulomatous inflammation in surgically removed material. Am J Surg Pathol. 1982;6(2):119-129.

(2.) El-Zammar OA, Katzenstein AL. Pathological diagnosis of granulomatous lung disease: a review. Histopathology. 2007;50(3):289-310.

(3.) Cheung OY, Muhm JR, Helmers RA, et al. Surgical pathology of granulomatous interstitial pneumonia. Ann Diagn Pathol. 2003;7(2):127-138.

(4.) Katzenstein A-L A. Katzenstein and Askin's Surgical Pathology of Non-Neoplastic Lung Disease. 4th ed. Philadelphia, PA: WB Saunders; 2006.

(5.) Adams DO. The granulomatous inflammatory response: a review. Am J Pathol. 1976;84(1):164-192.

(6.) Sarosi GA, Davies SF. Blastomycosis. Am Rev Respir Dis. 1979;120(4):911-938.

(7.) Schwarz J. The diagnosis of deep mycoses by morphologic methods. Hum Pathol. 1982;13(6):519-533.

(8.) Taxy JB. Blastomycosis: contributions of morphology to diagnosis--a surgical pathology, cytopathology, and autopsy pathology study. Am J Surg Pathol. 2007;31(4):615-623.

(9.) Tang YW, Procop GW, Zheng X, Myers JL, Roberts GD. Histologic parameters predictive of mycobacterial infection. Am J Clin Pathol. 1998;109(3): 331-334.

(10.) Ulbright TM, Katzenstein AL. Solitary necrotizing granulomas of the lung: differentiating features and etiology. Am J Surg Pathol. 1980;4(1):13-28.

(11.) Binford CH. Histoplasmosis: tissue reactions and morphologic variations of the fungus. Am J Clin Pathol. 1955;25(1):25-36.

(12.) Puckett TF. Pulmonary histoplasmosis; a study of twenty-two cases with identification of H. capsulatum in resected lesions. Am Rev Tuberc. 1953;67(4): 453-476.

(13.) Zimmerman LE. Demonstration of Histoplasma and Coccidioides in so-called tuberculomas of lung: preliminary report on thirty-five cases. AMA Arch Intern Med. 1954;94(5):690-699.

(14.) Greendyke RM, Emerson GL. Occurrence of histoplasma in solitary pulmonary nodules in a non-endemic area. Am J Clin Pathol. 1958;29(1):36-42.

(15.) Lazcano O, Speights VO Jr, Bilbao J, Becker J, Diaz J. Combined Fontana-Masson-mucin staining of Cryptococcus neoformans. Arch Pathol Lab Med. 1991;115(11):1145-1149.

(16.) Hayden RT, Qian X, Roberts GD, Lloyd RV. In situ hybridization for the identification of yeastlike organisms in tissue section. Diagn Mol Pathol. 2001; 10(1):15-23.

(17.) Koch ML, Cote RA. Comparison of fluorescence microscopy with ZiehlNeelsen stain for demonstration of acid-fast bacilli in smear preparations and tissue sections. Am Rev Respir Dis. 1965;91:283-284.

(18.) Kommareddi S, Abramowsky CR, Swinehart GL, Hrabak L. Nontuberculous mycobacterial infections: comparison of the fluorescent auramine-O and Ziehl-Neelsen techniques in tissue diagnosis. Hum Pathol. 1984;15(11):1085-1089.

(19.) Fukunaga H, Murakami T, Gondo T, Sugi K, Ishihara T. Sensitivity of acid-fast staining for Mycobacterium tuberculosis in formalin-fixed tissue. Am J Respir Crit Care Med. 2002;166(7):994-997.

(20.) Nayak NC, Sabharwal BD, Bhathena D, Mital GS, Ramalingaswami V. The pulmonary tuberculous lesion in North India: a study in medico-legal autopsies-I, incidence, nature, and evolution. Am Rev Respir Dis. 1970;101(1): 1-17.

(21.) Jannotta FS, Sidawy MK. The recognition of mycobacterial infections by intraoperative cytology in patients with acquired immunodeficiency syndrome. Arch Pathol Lab Med. 1989;113(10):1120-1123.

(22.) Smith MB, Molina CP, Schnadig VJ, Boyars MC, Aronson JF. Pathologic features of Mycobacterium kansasii infection in patients with acquired immunodeficiency syndrome. Arch Pathol Lab Med. 2003;127(5):554-560.

(23.) Snijder J. Histopathology of pulmonary lesions caused by atypical mycobacteria. J Pathol Bacteriol. 1965;90(1):65-73.

(24.) Attorri S, Dunbar S, Clarridge JE III. Assessment of morphology for rapid presumptive identification of Mycobacterium tuberculosis and Mycobacterium kansasii. J Clin Microbiol. 2000;38(4):1426-1429.

(25.) Gonzalez J, Tudo G, Gomez J, Garcia A, Navarro M, Jimenez de Anta MT. Use of microscopic morphology in smears prepared from radiometric cultures for presumptive identification of Mycobacterium tuberculosis complex, Mycobacterium avium complex, Mycobacterium kansasii, and Mycobacterium xenopi. Eur J Clin Microbiol Infect Dis. 1998;17(7):493-500.

(26.) Morris AJ, Reller LB. Reliability of cord formation in BACTEC media for presumptive identification of mycobacteria. J Clin Microbiol. 1993;31(9):2533-2534.

(27.) Farhi DC, Mason UG III, Horsburgh CR Jr. Pathologic findings in disseminated Mycobacterium avium-intracellulare infection: a report of 11 cases. Am J Clin Pathol. 1986;85(1):67-72.

(28.) Marchevsky A, Damsker B, Gribetz A, Tepper S, Geller SA. The spectrum of pathology of nontuberculous mycobacterial infections in open-lung biopsy specimens. Am J Clin Pathol. 1982;78(5):695-700.

(29.) Renshaw AA. The relative sensitivity of special stains and culture in open lung biopsies. Am J Clin Pathol. 1994;102(6):736-740.

(30.) Salian NV, Rish JA, Eisenach KD, Cave MD, Bates JH. Polymerase chain reaction to detect Mycobacterium tuberculosis in histologic specimens. Am J Respir Crit Care Med. 1998;158(4):1150-1155.

(31.) Schulz S, Cabras AD, Kremer M, et al. Species identification of mycobacteria in paraffin-embedded tissues: frequent detection of nontuberculous mycobacteria. Mod Pathol. 2005;18(2):274-282.

(32.) Goodwin RA Jr, Snell JD Jr. The enlarging histoplasmoma: concept of a tumor-like phenomenon encompassing the tuberculoma and coccidioidoma. Am Rev Respir Dis. 1969;100(1):1-12.

(33.) Medlar EM. The behavior of pulmonary tuberculous lesions; a pathological study. Am Rev Tuberc. 1955;71(3, pt 2):1-244.

(34.) Bhathena D, Mohapatra LN, Mital GS, Ramalingaswami V, Nayak NC. The pulmonary tuberculous lesion in North India: a study in medico-legal autopsies--II, bacteriologic aspects. Am Rev Respir Dis. 1970;101(1):18-26.

(35.) Myers JL, Katzenstein AL. Granulomatous infection mimicking bronchocentric granulomatosis. Am J Surg Pathol. 1986;10(5):317-322.

(36.) Takemura T, Matsui Y, Saiki S, Mikami R. Pulmonary vascular involvement in sarcoidosis: a report of 40 autopsy cases. Hum Pathol. 1992; 23(11):1216-1223.

(37.) Corpe RF, Stergus I. Is the histopathology of nonphotochromogenic mycobacterial infections distinguishable from that caused by Mycobacterium tuberculosis? Am Rev Respir Dis. 1963;87:289-291.

(38.) Merckx JJ, Soule EH, Karlson AG. The histopathology of lesions caused by infection with unclassified acid-fast bacteria in man: report of 25 cases. Am J Clin Pathol. 1964;41:244-255.

(39.) Griffith DE, Aksamit T, Brown-Elliott BA, et al. An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med. 2007;175(4):367-416.

(40.) Wallace JM, Hannah JB. Mycobacterium avium complex infection in patients with the acquired immunodeficiency syndrome: a clinicopathologic study. Chest. 1988;93(5):926-932.

(41.) Fujita J, Ohtsuki Y, Shigeto E, et al. Pathological findings of bronchiectases caused by Mycobacterium avium intracellulare complex. Respir Med. 2003; 97(8):933-938.

(42.) Fujita J, Ohtsuki Y, Suemitsu I, et al. Pathological and radiological changes in resected lung specimens in Mycobacterium avium intracellulare complex disease. Eur Respir J. 1999;13(3):535-540.

(43.) Kwon KY, Myers JL, Swensen SJ, Colby TV. Middle lobe syndrome: a clinicopathological study of 21 patients. Hum Pathol. 1995;26(3):302-307.

(44.) Bayer AS. Fungal pneumonias: pulmonary coccidioidal syndromes (part 2): miliary, nodular, and cavitary pulmonary coccidioidomycosis; chemotherapeutic and surgical considerations. Chest. 1981;79(6):686-691.

(45.) Deppisch LM, Donowho EM. Pulmonary coccidioidomycosis. Am J Clin Pathol. 1972;58(5):489-500.

(46.) DiCaudo DJ. Coccidioidomycosis: a review and update [quiz in J Am Acad Dermatol. 2006;55(6):943-945]. J Am Acad Dermatol. 2006;55(6):929-942.

(47.) Reynolds RJ III, Penn RL, Grafton WD, George RB. Tissue morphology of Histoplasma capsulatum in acute histoplasmosis. Am Rev Respir Dis. 1984; 130(2):317-320.

(48.) Spinello IM, Munoz A, Johnson RH. Pulmonary coccidioidomycosis. Semin Respir Crit Care Med. 2008;29(2):166-173.

(49.) Valdivia L, Nix D, Wright M, et al. Coccidioidomycosis as a common cause of community-acquired pneumonia. Emerg Infect Dis. 2006;12(6):958-962.

(50.) Wheat LJ, Conces D, Allen SD, Blue-Hnidy D, Loyd J. Pulmonary histoplasmosis syndromes: recognition, diagnosis, and management. Semin Respir Crit Care Med. 2004;25(2):129-144.

(51.) Chamany S, Mirza SA, Fleming JW, et al. A large histoplasmosis outbreak among high school students in Indiana, 2001. Pediatr Infect Dis J. 2004;23(10): 909-914.

(52.) Alznauer RL, Rolle C Jr, Pierce WF. Analysis of focalized pulmonary granulomas due to Coccidioides immitis. AMA Arch Pathol. 1955;59(6):641-655.

(53.) Huang CJ, Yang MC, Ueng SH. Large cryptococcoma mimicking lung cancer in an HIV-negative, type 2 diabetic patient. J Thorac Imaging. 2005;20(2): 115-117.

(54.) Goodwin RA Jr, Owens FT, Snell JD, et al. Chronic pulmonary histoplasmosis. Medicine (Baltimore). 1976;55(6):413-452.

(55.) Goodwin RA Jr, Shapiro JL, Thurman GH, Thurman SS, Des Prez RM. Disseminated histoplasmosis: clinical and pathologic correlations. Medicine (Baltimore). 1980;59(1):1-33.

(56.) Kauffman CA, Israel KS, Smith JW, White AC, Schwarz J, Brooks GF. Histoplasmosis in immunosuppressed patients. Am J Med. 1978;64(6):923-932.

(57.) Goodwin RA Jr, Snell JD, Hubbard WW, Terry RT. Early chronic pulmonary histoplasmosis. Am Rev Respir Dis. 1966;93(1):47-61.

(58.) Hutton JP, Durham JB, Miller DP, Everett ED. Hyphal forms of Histoplasma capsulatum: a common manifestation of intravascular infections. Arch Pathol Lab Med. 1985;109(4):330-332.

(59.) Weydert JA, Van Natta TL, DeYoung BR. Comparison of fungal culture versus surgical pathology examination in the detection of Histoplasma in surgically excised pulmonary granulomas. Arch Pathol Lab Med. 2007;131(5): 780-783.

(60.) Gal AA, Plummer AL, Langston AA, Mansour KA. Granulomatous Pneumocystis carinii pneumonia complicating hematopoietic cell transplantation [discussion in Pathol Res Pract. 2002;198(8):559-561]. Pathol Res Pract. 2002; 198(8):553-558.

(61.) Travis WD, Pittaluga S, Lipschik GY, et al. Atypical pathologic manifestations of Pneumocystis carinii pneumonia in the acquired immune deficiency syndrome: review of 123 lung biopsies from 76 patients with emphasis on cysts, vascular invasion, vasculitis, and granulomas. Am J Surg Pathol. 1990; 14(7):615-625.

(62.) Weber WR, Askin FB, Dehner LP. Lung biopsy in Pneumocystis carinii pneumonia: a histopathologic study of typical and atypical features. Am J Clin Pathol. 1977;67(1):11-19.

(63.) Baker RD, Haugen RK. Tissue changes and tissue diagnosis in cryptococcosis; a study of 26 cases. Am J Clin Pathol. 1955;25(1):14-24.

(64.) Haugen RK, Baker RD. The pulmonary lesions in cryptococcosis with special reference to subpleural nodules. Am J Clin Pathol. 1954;24(12):1381-1390.

(65.) McDonnell JM, Hutchins GM. Pulmonary cryptococcosis. Hum Pathol. 1985;16(2):121-128.

(66.) Gal AA, Koss MN, Hawkins J, Evans S, Einstein H. The pathology of pulmonary cryptococcal infections in the acquired immunodeficiency syndrome. Arch Pathol Lab Med. 1986;110(6):502-507.

(67.) Kent TH, Layton JM. Massive pulmonary crytococcosis. Am J Clin Pathol. 1962;38:596-604.

(68.) Farmer SG, Komorowski RA. Histologic response to capsule-deficient Cryptococcus neoformans. Arch Pathol. 1973;96(6):383-387.

(69.) Gal AA, Evans S, Meyer PR. The clinical laboratory evaluation of cryptococcal infections in the acquired immunodeficiency syndrome. Diagn Microbiol Infect Dis. 1987;7(4):249-254.

(70.) Ro JY, Lee SS, Ayala AG. Advantage of Fontana-Masson stain in capsule-deficient cryptococcal infection. Arch Pathol Lab Med. 1987;111(1):53-57.

(71.) Lombard CM, Tazelaar HD, Krasne DL. Pulmonary eosinophilia in coccidioidal infections. Chest. 1987;91(5):734-736.

(72.) Vanek J, Schwarz J, Hakim S. North American blastomycosis: a study of ten cases. Am J Clin Pathol. 1970;54(3):384-400.

(73.) Moore M. Morphologic variation in tissue of the organisms of the blastomycoses and of histoplasmosis. Am J Pathol. 1955;31(6):1049-1063.

(74.) Watts JC, Chandler FW, Mihalov ML, Kammeyer PL, Armin AR. Giant forms of Blastomyces dermatitidis in the pulmonary lesions of blastomycosis: potential confusion with Coccidioides immitis. Am J Clin Pathol. 1990;93(4): 575-578.

(75.) Cruickshank B. Pulmonary granulomatous pneumocystosis following renal transplantation: report of a case. Am J Clin Pathol. 1975;63(3):384-390.

(76.) Walsh TJ, Anaissie EJ, Denning DW, et al. Treatment of aspergillosis: clinical practice guidelines of the Infectious Diseases Society of America. Clin Infect Dis. 2008;46(3):327-360.

(77.) Binder RE, Faling LJ, Pugatch RD, Mahasaen C, Snider GL. Chronic necrotizing pulmonary aspergillosis: a discrete clinical entity. Medicine (Baltimore). 1982;61(2):109-124.

(78.) Yousem SA. The histological spectrum of chronic necrotizing forms of pulmonary aspergillosis. Hum Pathol. 1997;28(6):650-656.

(79.) Bosken CH, Myers JL, Greenberger PA, Katzenstein AL. Pathologic features of allergic bronchopulmonary aspergillosis. Am J Surg Pathol. 1988;12(3):216-222.

(80.) Katzenstein AL, Liebow AA, Friedman PJ. Bronchocentric granulomatosis, mucoid impaction, and hypersensitivity reactions to fungi. Am Rev Respir Dis. 1975;111(4):497-537.

(81.) Zander DS. Allergic bronchopulmonary aspergillosis: an overview. Arch Pathol Lab Med. 2005;129(7):924-928.

(82.) Travis WD, Kwon-Chung KJ, Kleiner DE, et al. Unusual aspects of allergic bronchopulmonary fungal disease: report of two cases due to Curvularia organisms associated with allergic fungal sinusitis. Hum Pathol. 1991;22(12): 1240-1248.

(83.) Irwin RS, Thomas HD. Mucoid inpaction of the bronchus: diagnosis and treatment. Am Rev Respir Dis. 1973;108(4):955-959.

(84.) Lam SM, Lau AC, Ma MW, Yam LY. Pseudallescheria boydii or Aspergillus fumigatus in a lady with an unresolving lung infiltrate, and a literature review. Respirology. 2008;13(3):478-480.

(85.) Sawada M, Isogai S, Miyake S, Kubota T, Yoshizawa Y. Pulmonary pseudallescherioma associated with systemic lupus erythematosus. Intern Med. 1998;37(12):1046-1049.

(86.) Asimacopoulos PJ, Katras A, Christie B. Pulmonary dirofilariasis: the largest single-hospital experience. Chest. 1992;102(3):851-855.

(87.) Ciferri F. Human pulmonary dirofilariasis in the United States: a critical review. Am J Trop Med Hyg. 1982;31(2):302-308.

(88.) Ro JY, Tsakalakis PJ, White VA, et al. Pulmonary dirofilariasis: the great imitator of primary or metastatic lung tumor--a clinicopathologic analysis of seven cases and a review of the literature. Hum Pathol. 1989;20(1):69-76.

(89.) Flieder DB, Moran CA. Pulmonary dirofilariasis: a clinicopathologic study of 41 lesions in 39 patients. Hum Pathol. 1999;30(3):251-256.

(90.) Gutierrez Y. Diagnostic features of zoonotic filariae in tissue sections. Hum Pathol. 1984;15(6):514-525.

(91.) Gal AA, Koss MN. The pathology of sarcoidosis. Curr Opin Pulm Med. 2002;8(5):445-451.

(92.) Carrington CB. Structure and function in sarcoidosis. Ann N Y Acad Sci. 1976;278:265-283.

(93.) Iwai K, Oka H. Sarcoidosis: report of ten autopsy cases in Japan. Am Rev Respir Dis. 1964;90:612-622.

(94.) Longcope WT, Freiman DG. A study of sarcoidosis; based on a combined investigation of 160 cases including 30 autopsies from The Johns Hopkins Hospital and Massachusetts General Hospital. Medicine (Baltimore). 1952;31(1): 1-132.

(95.) Lynch JP III, Kazerooni EA, Gay SE. Pulmonary sarcoidosis. Clin Chest Med. 1997;18(4):755-785.

(96.) Reid JD, Andersen ME. Calcium oxalate in sarcoid granulomas: with particular reference to the small ovoid body and a note on the finding of dolomite. Am J Clin Pathol. 1988;90(5):545-558.

(97.) Visscher D, Churg A, Katzenstein AL. Significance of crystalline inclusions in lung granulomas. Mod Pathol. 1988;1(6):415-419.

(98.) Rosen Y, Moon S, Huang CT, Gourin A, Lyons HA. Granulomatous pulmonary angiitis in sarcoidosis. Arch Pathol Lab Med. 1977;101(4):170-174.

(99.) Baum GL, Schwarz J, Barlow PB. Sarcoidosis and specific etiologic agents: a continuing enigma. Chest. 1973;63(4):488-494.

(100.) Churg AM, Green FHY. Occupational lung disease. In: Churg AM, Myers JL, Tazelaar HD, Wright JL, eds. Thurlbeck's Pathology of the Lung. 3rd ed. New York, NY: Thieme; 2005:769-862.

(101.) Freiman DG, Hardy HL. Beryllium disease: the relation of pulmonary pathology to clinical course and prognosis based on a study of 130 cases from the U.S. beryllium case registry. Hum Pathol. 1970;1(1):25-44.

(102.) Williams WJ. A histological study of the lungs in 52 cases of chronic beryllium disease. Br J Ind Med. 1958;15(2):84-91.

(103.) Frome EL, Newman LS, Cragle DL, Colyer SP, Wambach PF. Identification of an abnormal beryllium lymphocyte proliferation test. Toxicology. 2003;183(1-3):39-56.

(104.) Maier LA, Martyny JW, Liang J, Rossman MD. Recent chronic beryllium disease in residents surrounding a beryllium facility. Am J Respir Crit Care Med. 2008;177(9):1012-1017.

(105.) Coleman A, Colby TV. Histologic diagnosis of extrinsic allergic alveolitis. Am J Surg Pathol. 1988;12(7):514-518.

(106.) Emanuel DA, Wenzel FJ, Bowerman CI, Lawton BR. Farmer's lung: clinical, pathologic and immunologic study of twenty-four patients. Am J Med. 1964;37:392-401.

(107.) Reyes CN, Wenzel FJ, Lawton BR, Emanuel DA. The pulmonary pathology of farmer's lung disease. Chest. 1982;81(2):142-146.

(108.) Barrios RJ. Hypersensitivity pneumonitis: histopathology. Arch Pathol Lab Med. 2008;132(2):199-203.

(109.) Agarwal R, Nath A. Hot-tub lung: hypersensitivity to Mycobacterium avium but not hypersensitivity pneumonitis. Respir Med. 2006;100(8):1478.

(110.) Aksamit TR. Hot tub lung: infection, inflammation, or both? Semin Respir Infect. 2003;18(1):33-39.

(111.) Cappelluti E, Fraire AE, Schaefer OP. A case of "hot tub lung" due to Mycobacterium avium complex in an immunocompetent host. Arch Intern Med. 2003;163(7):845-848.

(112.) Embil J, Warren P, Yakrus M, et al. Pulmonary illness associated with exposure to Mycobacterium-avium complex in hot tub water: hypersensitivity pneumonitis or infection? Chest. 1997;111(3):813-816.

(113.) Grimes MM, Cole TJ, Fowler AA III. Obstructive granulomatous bronchiolitis due to Mycobacterium avium complex in an immuno-competent man. Respiration. 2001;68(4):411-415.

(114.) Hanak V, Kalra S, Aksamit TR, Hartman TE, Tazelaar HD, Ryu JH. Hot tub lung: presenting features and clinical course of 21 patients. Respir Med. 2006; 100(4):610-615.

(115.) Hartman TE, Jensen E, Tazelaar HD, Hanak V, Ryu JH. CT findings of granulomatous pneumonitis secondary to Mycobacterium avium-intracellulare inhalation: "hot tub lung". AJR Am J Roentgenol. 2007;188(4):1050-1053.

(116.) Kahana LM, Kay JM, Yakrus MA, Waserman S. Mycobacterium avium complex infection in an immunocompetent young adult related to hot tub exposure. Chest. 1997;111(1):242-245.

(117.) Khoor A, Leslie KO, Tazelaar HD, Helmers RA, Colby TV. Diffuse pulmonary disease caused by nontuberculous mycobacteria in immunocompetent people (hot tub lung). Am J Clin Pathol. 2001;115(5):755-762.

(118.) Mangione EJ, Huitt G, Lenaway D, et al. Nontuberculous mycobacterial disease following hot tub exposure. Emerg Infect Dis. 2001;7(6):1039-1042.

(119.) Marchetti N, Criner K, Criner GJ. Characterization of functional, radiologic and lung function recovery post-treatment of hot tub lung: a case report and review of the literature. Lung. 2004;182(5):271-277.

(120.) Pham RV, Vydareny KH, Gal AA. High-resolution computed tomography appearance of pulmonary Mycobacterium avium complex infection after exposure to hot tub: case of hot-tub lung. J Thorac Imaging. 2003;18(1):48-52.

(121.) Rickman OB, Ryu JH, Fidler ME, Kalra S. Hypersensitivity pneumonitis associated with Mycobacterium avium complex and hot tub use. Mayo Clin Proc. 2002;77(11):1233-1237.

(122.) Sood A, Sreedhar R, Kulkarni P, Nawoor AR. Hypersensitivity pneumonitis-like granulomatous lung disease with nontuberculous mycobacteria from exposure to hot water aerosols. Environ Health Perspect. 2007;115(2):262-266.

(123.) Travaline JM, Kelsen SG. Hypersensitivity pneumonitis associated with hot tub use [reply in Arch Intern Med. 2003;163(18):2250-2251]. Arch Intern Med. 2003;163(18):2250.

(124.) Marras TK, Wallace RJ Jr, Koth LL, Stulbarg MS, Cowl CT, Daley CL. Hypersensitivity pneumonitis reaction to Mycobacterium avium in household water. Chest. 2005;127(2):664-671.

(125.) Glazer CS, Martyny JW, Lee B, et al. Nontuberculous mycobacteria in aerosol droplets and bulk water samples from therapy pools and hot tubs. J Occup Environ Hyg. 2007;4(11):831-840.

(126.) Cha SI, Fessler MB, Cool CD, Schwarz MI, Brown KK. Lymphoid interstitial pneumonia: clinical features, associations and prognosis. Eur Respir J. 2006;28(2):364-369.

(127.) Swigris JJ, Berry GJ, Raffin TA, Kuschner WG. Lymphoid interstitial pneumonia: a narrative review. Chest. 2002;122(6):2150-2164.

(128.) Teruya-Feldstein J, Kingma DW, Weiss A, et al. Chemokine gene expression and clonal analysis of B cells in tissues involved by lymphoid interstitial pneumonitis from HIV-infected pediatric patients. Mod Pathol. 2001; 14(10):929-936.

(129.) Koss MN, Hochholzer L, Langloss JM, Wehunt WD, Lazarus AA. Lymphoid interstitial pneumonia: clinicopathological and immunopathological findings in 18 cases. Pathology. 1987;19(2):178-185.

(130.) Jennette JC, Falk RJ. Small-vessel vasculitis. N Engl J Med. 1997;337(21): 1512-1523.

(131.) Finkielman JD, Lee AS, Hummel AM, et al. ANCA are detectable in nearly all patients with active severe Wegener's granulomatosis. Am J Med. 2007; 120(7):643.e9-643.e14.

(132.) Mark EJ, Matsubara O, Tan-Liu NS, Fienberg R. The pulmonary biopsy in the early diagnosis of Wegener's (pathergic) granulomatosis: a study based on 35 open lung biopsies. Hum Pathol. 1988;19(9):1065-1071.

(133.) Carrington CB, Liebow A. Limited forms of angiitis and granulomatosis of Wegener's type. Am J Med. 1966;41(4):497-527.

(134.) Katzenstein AL, Locke WK. Solitary lung lesions in Wegener's granulomatosis: pathologic findings and clinical significance in 25 cases. Am J Surg Pathol. 1995;19(5):545-552.

(135.) Liebow AA. The J. Burns Amberson lecture--pulmonary angiitis and granulomatosis. Am Rev Respir Dis. 1973;108(1):1-18.

(136.) Travis WD, Hoffman GS, Leavitt RY, Pass HI, Fauci AS. Surgical pathology of the lung in Wegener's granulomatosis: review of 87 open lung biopsies from 67 patients. Am J Surg Pathol. 1991;15(4):315-333.

(137.) Uner AH, Rozum-Slota B, Katzenstein AL. Bronchiolitis obliteransorganizing pneumonia (BOOP)-like variant of Wegener's granulomatosis: a clinicopathologic study of 16 cases. Am J Surg Pathol. 1996;20(7):794-801.

(138.) Fienberg R, Mark EJ, Goodman M, McCluskey RT, Niles JL. Correlation of antineutrophil cytoplasmic antibodies with the extrarenal histopathology of Wegener's (pathergic) granulomatosis and related forms of vasculitis. Hum Pathol. 1993;24(2):160-168.

(139.) Gal AA, Velasquez A. Antineutrophil cytoplasmic autoantibody in the absence of Wegener's granulomatosis or microscopic polyangiitis: implications for the surgical pathologist. Mod Pathol. 2002;15(3):197-204.

(140.) Lombard CM, Duncan SR, Rizk NW, Colby TV. The diagnosis of Wegener's granulomatosis from transbronchial biopsy specimens. Hum Pathol. 1990;21(8):838-842.

(141.) Yousem SA, Lombard CM. The eosinophilic variant of Wegener's granulomatosis. Hum Pathol. 1988;19(6):682-688.

(142.) Yoshikawa Y, Watanabe T. Pulmonary lesions in Wegener's granulomatosis: a clinicopathologic study of 22 autopsy cases. Hum Pathol. 1986;17(4): 401-410.

(143.) Gaudin PB, Askin FB, Falk RJ, Jennette JC. The pathologic spectrum of pulmonary lesions in patients with anti-neutrophil cytoplasmic autoantibodies specific for anti-proteinase 3 and anti-myeloperoxidase. Am J Clin Pathol. 1995; 104(1):7-16.

(144.) Myers JL, Katzenstein AL. Wegener's granulomatosis presenting with massive pulmonary hemorrhage and capillaritis. Am J Surg Pathol. 1987;11(11): 895-898.

(145.) Travis WD, Carpenter HA, Lie JT. Diffuse pulmonary hemorrhage: an uncommon manifestation of Wegener's granulomatosis. Am J Surg Pathol. 1987; 11(9):702-708.

(146.) Yousem SA. Bronchocentric injury in Wegener's granulomatosis: a report of five cases. Hum Pathol. 1991;22(6):535-540.

(147.) Travis WD, Colby TV, Lombard C, Carpenter HA. A clinicopathologic study of 34 cases of diffuse pulmonary hemorrhage with lung biopsy confirmation. Am J Surg Pathol. 1990;14(12):1112-1125.

(148.) Churg A. Pulmonary angiitis and granulomatosis revisited. Hum Pathol. 1983;14(10):868-883.

(149.) Katzenstein AL. Necrotizing granulomas of the lung. Hum Pathol. 1980; 11(6):596-597.

(150.) Churg J, Strauss L. Allergic granulomatosis, allergic angiitis, and periarteritis nodosa. Am J Pathol. 1951;27(2):277-301.

(151.) Churg A. Recent advances in the diagnosis of Churg-Strauss syndrome. Mod Pathol. 2001;14(12):1284-1293.

(152.) Katzenstein AL. Diagnostic features and differential diagnosis of Churg-Strauss syndrome in the lung: a review. Am J Clin Pathol. 2000;114(5):767-772.

(153.) Koss MN, Antonovych T, Hochholzer L. Allergic granulomatosis (Churg-Strauss syndrome): pulmonary and renal morphologic findings. Am J Surg Pathol. 1981;5(1):21-28.

(154.) Reid AJ, Harrison BD, Watts RA, Watkin SW, McCann BG, Scott DG. Churg-Strauss syndrome in a district hospital. QJM. 1998;91(3):219-229.

(155.) Marom EM, McAdams HP, Erasmus JJ, Goodman PC. The many faces of pulmonary aspiration. AJR Am J Roentgenol. 1999;172(1):121-128.

(156.) Mukhopadhyay S, Katzenstein AL. Pulmonary disease due to aspiration of food and other particulate matter: a clinicopathologic study of 59 cases diagnosed on biopsy or resection specimens. Am J Surg Pathol. 2007;31(5): 752-759.

(157.) Crome L, Valentine JC. Pulmonary nodular granulomatosis caused by inhaled vegetable particles. J Clin Pathol. 1962;15:21-25.

(158.) Knoblich R. Pulmonary granulomatosis caused by vegetable particles: so-called lentil pulse pneumonia. Am Rev Respir Dis. 1969;99(3):380-389.

(159.) Mitchinson MJ. The inhalation of leguminous particles. J Pathol. 1972; 107(3):221-222.

(160.) Moran TJ. Experimental food-aspiration pneumonia. AMA Arch Pathol. 1951;52(4):350-354.

(161.) Teabeaut JRII. Aspiration of gastric contents; an experimental study. Am J Pathol. 1952;28(1):51-67.

(162.) Thompson JR. Identification of foreign material in the lungs. Am Rev Respir Dis. 1962;85:281.

(163.) Barnes TW, Vassallo R, Tazelaar HD, Hartman TE, Ryu JH. Diffuse bronchiolar disease due to chronic occult aspiration. Mayo Clin Proc. 2006; 81(2):172-176.

(164.) Bulmer SR, Lamb D, McCormack RJ, Walbaum PR. Aetiology of unresolved pneumonia. Thorax. 1978;33(3):307-314.

(165.) Emery JL. Two cases of lentil pneumonitis. Proc R Soc Med. 1960;53: 952-953.

(166.) Fetterman GH, Moran TJ. Food aspiration pneumonia. Penn Med J. 1942; 45:810-812.

(167.) Head MA. Foreign body reaction to inhalation of lentil soup: giant cell pneumonia. J Clin Pathol. 1956;9(4):295-299.

(168.) Kaplan SL, Gnepp DR, Katzenstein AL, Feigin RD. Miliary pulmonary nodules due to aspirated vegetable particles. J Pediatr. 1978;92(3):448-450.

(169.) Marom EM, McAdams HP, Sporn TA, Goodman PC. Lentil aspiration pneumonia: radiographic and CT findings. J Comput Assist Tomogr. 1998;22(4): 598-600.

(170.) Ros PR. Lentil aspiration pneumonia. JAMA. 1984;251(10):1277-1278.

(171.) Vidyarthi SC. Diffuse miliary granulomatosis of the lungs due to aspirated vegetable cells. Arch Pathol. 1967;83(2):215-218.

(172.) Chaplin AJ, Kemp P. Polystyrene sulphonates and aspiration pneumonia: a histochemical study. Histopathology. 1984;8(2):331-338.

(173.) Haupt HM, Hutchins GM. Sodium polystyrene sulfonate pneumonitis. Arch Intern Med. 1982;142(2):379-381.

(174.) Oi RH. The microscopic appearance of a sodium-potassium exchange resin in histologic sections. Am J Clin Pathol. 1978;69(3):359-361.

(175.) Kahn J. A cloud of smoke: the complicated death of a 9/11 hero. The New Yorker. September 15, 2008:6.

(176.) Douglas FG, Kafilmout KJ, Patt NL. Foreign particle embolism in drug addicts: respiratory pathophysiology. Ann Intern Med. 1971;75(6):865-880.

(177.) Ganesan S, Felo J, Saldana M, Kalasinsky VF, Lewin-Smith MR, Tomashefski JF Jr. Embolized crospovidone (poly[N-vinyl-2-pyrrolidone]) in the lungs of intravenous drug users. Mod Pathol. 2003;16(4):286-292.

(178.) Kringsholm B, Christoffersen P. The nature and the occurrence of birefringent material in different organs in fatal drug addiction. Forensic Sci Int. 1987;34(1-2):53-62.

(179.) Lamb D, Roberts G. Starch and talc emboli in drug addicts' lungs. J Clin Pathol. 1972;25(10):876-881.

(180.) Pare JA, Fraser RG, Hogg JC, Howlett JG, Murphy SB. Pulmonary 'mainline' granulomatosis: talcosis of intravenous methadone abuse. Medicine (Baltimore). 1979;58(3):229-239.

(181.) Tomashefski JF Jr, Felo JA. The pulmonary pathology of illicit drug and substance abuse. Curr Diagn Pathol. 2004;10:413-426.

(182.) Tomashefski JF Jr, Hirsch CS. The pulmonary vascular lesions of intravenous drug abuse. Hum Pathol. 1980;11(2):133-145.

(183.) Tomashefski JF Jr, Hirsch CS, Jolly PN. Microcrystalline cellulose pulmonary embolism and granulomatosis: a complication of illicit intravenous injections of pentazocine tablets. Arch Pathol Lab Med. 1981; 105(2):89-93.

(184.) Lewman LV. Fatal pulmonary hypertension from intravenous injection of methylphenidate (Ritalin) tablets. Hum Pathol. 1972;3(1):67-70.

(185.) Houck RJ, Bailey GL, Daroca PJ Jr, Brazda F, Johnson FB, Klein RC. Pentazocine abuse: report of a case with pulmonary arterial cellulose granulomas and pulmonary hypertension. Chest. 1980;77(2):227-230.

(186.) Zeltner TB, Nussbaumer U, Rudin O, Zimmermann A. Unusual pulmonary vascular lesions after intravenous injections of microcrystalline cellulose: a complication of pentazocine tablet abuse. Virchows Arch A Pathol Anat Histol. 1982;395(2):207-216.

(187.) Szwed JJ. Pulmonary angiothrombosis caused by "blue velvet" addiction. Ann Intern Med. 1970;73(5):771-774.

(188.) Kleinfeld M, Giel CP, Majeranowski JF, Messite J. Talc pneumoconiosis: a report of six patients with postmortem findings. Arch Environ Health. 1963;7: 101-115.

(189.) Vallyathan NV, Craighead JE. Pulmonary pathology in workers exposed to nonasbestiform talc. Hum Pathol. 1981;12(1):28-35.

(190.) Abraham JL, Brambilla C. Particle size for differentiation between inhalation and injection pulmonary talcosis. Environ Res. 1980;21(1):94-96.

(191.) Baruch AC, Steinbronn K, Sobonya R. Pulmonary adenocarcinomas associated with rheumatoid nodules: a case report and review of the literature. Arch Pathol Lab Med. 2005;129(1):104-106.

(192.) Beumer HM, van Belle CJ. Pulmonary nodules in rheumatoid arthritis. Respiration. 1972;29(5):556-564.

(193.) Yousem SA, Colby TV, Carrington CB. Lung biopsy in rheumatoid arthritis. Am Rev Respir Dis. 1985;131(5):770-777.

(194.) Koss MN, Robinson RG, Hochholzer L. Bronchocentric granulomatosis. Hum Pathol. 1981;12(7):632-638.

(195.) Myers JL. Bronchocentric granulomatosis: disease or diagnosis? Chest. 1989;96(1):3-4.

(196.) Tazelaar HD, Baird AM, Mill M, Grimes MM, Schulman LL, Smith CR. Bronchocentric mycosis occurring in transplant recipients. Chest. 1989;96(1):92-95.

Sanjay Mukhopadhyay, MD; Anthony A. Gal, MD

Accepted for publication April 16, 2009.

From the Department of Pathology, State University of New York Upstate Medical University, Syracuse, New York (Dr Mukhopadhyay); and the Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, Georgia (Dr Gal).

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Sanjay Mukhopadhyay, MD, Department of Pathology, State University of New York Upstate Medical University, 750 East Adams Street, Syracuse, NY 13210 (e-mail: mukhopas@upstate.edu).
Table 1. Differential Diagnosis of Granulomatous
Lung Disease

Infections
  Mycobacteria
    Mycobacterium tuberculosis
    Nontuberculous mycobacteria
  Fungi
    Histoplasma
    Cryptococcus
    Coccidioides
    Blastomyces
    Pneumocystis
    Aspergillus
  Parasites
    Dirofilaria
Noninfectious diseases
  Sarcoidosis
  Chronic beryllium disease
  Hypersensitivity pneumonitis
  Hot tub lung
  Lymphoid interstitial pneumonia
  Wegener granulomatosis
  Churg-Strauss syndrome
  Aspiration pneumonia
  Talc granulomatosis
  Rheumatoid nodule
  Bronchocentric granulomatosis

Table 2. Key Diagnostic Features of Major Noninfectious Granulomatous
Lung Diseases

Key Features                                      Diagnosis

Prominent, well-formed, discrete,                 Sarcoidosis
nonnecrotizing granulomas in pleura,
interlobular septa, and walls of bronchioles
(a)
Normal lung away from granulomas

Prominent interstitial chronic inflammation       Hypersensitivity
Scattered, small, poorly formed granulomas or     pneumonitis
multinucleated giant cells in interstitium

Granulomas within bronchiolar lumens              Hot tub lung
History of hot tub use (b)

Suppurative granulomas with "dirty" necrosis      Wegener
Necrotizing vasculitis                            granulomatosis

Necrotizing granulomas                            Churg-Strauss
Necrotizing vasculitis                            syndrome
Prominent eosinophils

Vegetable material surrounded by foreign          Aspiration pneumonia
body-type granulomas or multinucleated giant
cells

Interstitial foreign body-type granulomas         Talc granulomatosis
containing talc, microcrystalline cellulose,
or crospovidone

Active, seropositive rheumatoid arthritis (b)     Rheumatoid nodule
Multiple, bilateral lung nodules
Subpleural necrotizing granuloma

(a) Features not consistent with sarcoidosis include extensive
necrosis or suppuration, interstitial inflammation away from the
granulomas, organizing pneumonia, granulomas within alveolar or
bronchiolar airspaces, numerous eosinophils, and vegetable material.

(b) Clinical information essential for diagnosis.

Table 3. Morphologic Features of the Common Fungi Causing
Granulomatous Lung Disease

                               Histoplasma              Cryptococcus

Visualized on         No (in necrotizing granulomas)    Yes
  hematoxylineosin    Yes (in disseminated
                        histoplasmosis)
Spherules and         No                                No
  endospores
Usual sizea           Small                             Small
                      3 mm                              4-7 [micro]m
Shape                 Mostly oval, often tapered        Round

                        at one or both ends
Size variation        Minimal                           Marked
Budding               Narrow based                      Occasional
Nuclei                Single (when intracellular)       None
Staining with         Absent                            Usual, strong
  mucicarmine

                          Coccidioides        Blastomyces

Visualized on         Yes                     Yes
  hematoxylineosin

Spherules and         Yes                     No
  endospores
Usual sizea           Large                   Large
                      30-60 mm (spherules)    8-15 [micro]m
Shape                 2-5 mm (endospores)     Round
                      Round or fragmented
                        (spherules)
                      Round (endospores)
Size variation        Marked                  Considerable
Budding               Absent                  Broad based
Nuclei                None                    Multiple
Staining with         Absent                  Occasional,
  mucicarmine                                   weak

(a) Sizes listed are the usual sizes encountered. Greater size
variation has been reported with most of the fungi listed2,4,7 but is
unusual.
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.