146

Granulomatous Infections: Etiology and Classification

Alimuddin Zumla and D. Geraint James From the Academic Infectious Diseases Unit, Department of Medicine,
University College London Medical School; and the Royal Free
Hospital School of Medicine, London, United Kingdom

Granulomatous disorders are frequently due to a wide variety of infections. Over the past decade
advances in molecular diagnostic techniques have allowed identification of organisms involved in
granulomatous disorders that previously were of unknown etiol~gy. ~n the basis of cu~ently avail-
able information, granulomatous infections can now be classified m three categones. Group 1
infections are due to a well-recognized organism. Group 2 comprises infections due to organisms
that have been recently identified in granulomas by molecular methods but are not readily isolated
by conventional microbiological techniques. Group 3 consists of disorders for which the causal
organisms have not yet been identified but are strongly suspected; further a~van~es in di~gnostic
techniques will lead to reclassification of some of these disorders as grou~ 2. ThIs r~vlew descnbe~ the
etiology,histopathologic features, and classification of granulomatous disorders, with an emphasis on
those of groups 2 and 3.

Granuloma Formation patibility (MHC) class II molecules. The T helper cells recog-
nize protein peptides presented to it by antigen-presenting cells
A granuloma can be defined as a focal, compact collection
bearing MHC class II molecules.
of inflammatory cells in which mononuclear cells predominate.
The T cell induces IL-l on the macrophage, and thereafter
Granulomas usually form as a result of the persistence of a
a cavalcade of chemotactic factors promote granuloma genesis
nondegradable product or as the result of hypersensitivity re-
[2, 4]. IFN-'}' increases the expression of MHC class II mole-
sponses [1]. An overlap of the two mechanisms occurs in most
cules on macrophages, and activated macrophage receptors
infectious diseases since microorganisms can serve both as
carry a crystallizable fraction of IgG to potentiate their ability
foreign bodies and as antigens for immunologic responses.
to phagocytose [4]. The end result is the epithelioid granuloma,
Normally granulomas are the result of protective mechanisms
which progresses toward fibrosis under the impact of trans-
and form when acute inflammatory processes cannot destroy
forming growth factor and platelet-derived growth factor.
invading agents.
T cell subsets, macrophages, and other immune cells produce
The granuloma forms by a stepwise series of events and is
several cytokines, the amounts and types of which determine
the end result ofa complex interplay among invading organism,
further subclassifications of granulomas. CD4 + T helper (Th)
antigen, chemical, drug or other irritant, prolonged antigene-
lymphocytes are necessary for the development of granulomas
mia, macrophage activity, T cell responses, B cell overactivity,
and are broadly classified into two functional types, Thl and
circulating immune complexes, and a vast array of biological
Th2 [5]. Thl-type cells produce IL-2, IFN-'}', and TNF-,B upon
mediators [2]. Areas ofinflammation or immunologic reactivity
stimulation with antigen and also participate in delayed-type
attract monocyte-macrophages, which may fuse to form multi-
hypersensitivity responses, while Th2-type cells make IL-4,
nucleated giant cells [3]. Further cellular transformation of
IL-5, IL-6, IL-9, IL-I0, and IL-13, cytokines which are im-
macrophages to epithelioid cells may occur.
portant for development of B cells and eosinophilia.
The granuloma is an active site of numerous enzymes and
There appears to be a complex relationship between Thl-
cytokines and, with aging, fibronectin and progression factors
type and Th2-type responses and infectious granulomas. Gran-
such as platelet-derived growth factor, transforming growth
ulomas that synthesize predominantly Th2-type cytokines, such
factor-,B, insulin-like growth factor, and TNF-a. There is a
as those that form in response to parasite ova, make only small
close relationship between CD4 + T lymphocytes and activated
quantities of Thl-type cytokines chronically [6], whereas gran-
macrophages showing increased expression of major histocom-
ulomas such as those of tuberculoid leprosy make large quanti-
ties of Thl-type cytokines and less of Th2-type lymphokines
[7]. Granulomas of various infections may have different im-
Received 21 August 1995; revised 23 February 1996.
munoregulatory mechanisms governing their formation and
Financial support: Camden and Islington Community Health Services NHS resolution [I, 2, 7].
Trust.
Reprints or correspondence: Professor D. Geraint James, Visiting ~rofessor
of Medicine, Royal Free Hospital School of Medicine, Rowland HIll Street,
London NW3 2PF, United Kingdom Classification of Granulomatous Infections
Clinical Infectious Diseases 1996;23:146-58
1996 by The University of Chicago. All rights reserved.
Granulomatous disorders are frequently due to a wide variety
1058-4838/96/2301-0020$02.00 of infections. Over the past decade advances in molecular diag-
em 1996;23 (July) Granulomatous Infections 147

Table 1. Classification of granulomatous infections and associated Group 1 Granulomatous Infections (Causal Agents
pathogens. Well Defined)
Group number, A wide range of infectious organisms cause the granuloma-
type of causal agents Granulomatous disorders
tous diseases that fall into this category (table 1). These diseases
Group I
share similar histologic features and have identifiable and dis-
Well recognized tinct etiologic agents.
Mycobacteria Tuberculosis, leprosy, Buruli ulcer,
swimming pool (fish tank)
granuloma Mycobacterial Infections
Bacteria Brucellosis, melioidosis, actinomycosis,
nocardiosis, granuloma inguinale, Mycobacteria are a large group comprising Mycobacterium
listeriosis, tularemia tuberculosis, Mycobacterium leprae, Mycobacterium avium
Spirochetes Syphilis, pinta, yaws
complex, Mycobacterium ulcerans, Mycobacterium marinum.
Fungi Mycoses (see table 3)
Protozoa Leishmaniasis, toxoplasmosis several opportunistic mycobacteria, and the mycobacteria from
Nematodes Visceral larva migrans (toxocariasis) which BCG vaccine is produced. The nontuberculous myco-
Trematodes Schistosomiasis, paragonimiasis, bacteria produce a wide clinical spectrum of granulomatous
fascioliasis, clonorchiasis disorders, summarized in table 2. Mycobacteria have been im-
Chlamydia species Lymphogranuloma venereum, trachoma
plicated in the etiopathogenesis of sarcoidosis and Crohn' s
Rickettsia Q fever (Coxiella burnetii infection)
Viruses Infectious mononucleosis, disease, and this is discussed in the section on group 3 disor-
cytomegalovirus infection, measles, ders.
mumps Tuberculosis. The etiologic agent of tuberculosis, M. tu-
Group 2 berculosis, is clinically one of the most common microbial
Recently recognized
causes of granulomas [2, 8]. The M. tuberculosis granuloma,
Bacterium Cat-scratch disease (Bartonella
henselae infection) classically characterized by the presence of central caseous
Actinomyces Whipple's disease (Tropheryma necrosis, is referred to as a tubercle. Typically, there is a central
whippelii infection) area of amorphous caseating granular debris and loss of cellular
Group 3 detail, and acid-fast bacilli are present. This area is encircled
Suspected but not established
by epithelioid cells, lymphocytes, histiocytes, fibroblasts, and
Measles virus Crohn's disease
Mycobacterium Primary biliary cirrhosis occasionally Langhans' giant cells. These histologic features
? (see table 4) Sarcoidosis of the granuloma are sufficiently characteristic to allow reason-
Viral Kikuchi's disease ably accurate diagnosis of tuberculosis [8].
? Chronic granulomatous disease of While caseating granulomas are the classic finding in cases
childhood
of tuberculosis, they are not always present; noncaseating gran-
Langerhans granulomatosis
ulomas may occur. When acid-fast bacilli are not identifiable
in the granuloma, culture of the biopsy specimen may yield
M. tuberculosis. PCR with use of primers specific for M. tuber-
culosis is currently under evaluation for the identification of
nostic techniques have allowed identification of organisms this species [9] and may develop into an important tool for
involved in granulomatous disorders that previously were of distinguishing it from other (nontuberculous) mycobacteria.
unknown etiology. On the basis of currently available informa- Immunosuppression may alter the classic histopathologic
tion, granulomas associated with infections can be categorized features of tuberculosis. A spectrum of histopathologic granu-
in three groups, as listed in table 1. lomatous tissue reactions are seen in patients infected with HIV
Group 1 granulomatous disorders are caused by well-recog- [10]. Classic well-formed granulomas have been observed in
nized agents. In group 2, the causal agents have been suspected individuals with early HIV disease whose CD4 cell counts have
over the years but have only recently been recognized by newer remained adequate. In patients with advanced HIV disease, the
diagnostic molecular techniques such as PCR. These agents are granulomas are less well formed, are more necrotic, and may
not readily isolated by conventional bacteriologic techniques. contain abundant bacilli, which may be demonstrated by appro-
Group 3 comprises granulomatous disorders that may be associ- priate staining techniques as well as by culture of biopsy speci-
ated with particular infectious agents but for which the causal mens [9-11].
organisms have not yet been determined. Future advances in Leprosy. Leprosy, or Hansen's disease, is a chronic granu-
diagnostic techniques may provide strong evidence of the lomatous disorder caused by M. leprae, an obligate intracellular
causal agents, in which case the disorders may be reclassified bacterium that predominantly affects the skin and nerves [12].
as group 2; any such advancements will allow better manage- First identified over a century ago by Armauer Hansen, the
ment of these idiopathic granulomatoses. bacterium has defied all attempts to grow it in culture. Leprosy
148 Zumla and James
Table 2. Granulomatous diseases caused by nontuberculous mycobacteria.
Clinical manifestation
Site or type of disease
Lung Bronchopulmonary disease
Lymph node Lymphadenitis
Bones and joints Osteomyelitis
Skin and soft tissue Abscesses
Buruli ulcer
Swimming pool granuloma
Tuberculoid leprosy
Multisystem Lepromatous leprosy
involvement
Meningitis, diarrhea, pneumonia,
lymphadenitis
manifests as a spectrum of well-described clinical, pathologi-
cal, and immunologic features [12].
At one end of the spectrum is lepromatous leprosy, the low
resistance form with multiple lesions, extensive skin and vis-
ceral involvement, and diffuse infiltration of lesions with nu-
merous leprotic baccilli. At the other end of the spectrum,
tuberculoid leprosy, the highly resistant form, is characterized
by one or few skin lesions and involvement of the nerve at the
site of the lesion. Skin biopsies of patients with tuberculoid
leprosy reveal a few acid-fast and alcohol-fast leprotic bacilli
and a well-defined giant cell granuloma. Actual granulomatous
invasion and destruction of dermal nerves, occasionally with
caseous necrosis, sometimes occurs.
Within the dermis of patients with lepromatous leprosy, in-
creased levels of Th2-type cytokines (lL-4, IL-5, and IL-IO)
and decreased levels of IL-2 and IFN-y have been noted. In
contrast, Th I-type cytokines predominate in tuberculoid lesions
[7]. The diagnosis of leprosy is based on a combination of
clinical and histologic findings. The paucity of acid-fast bacilli
in the lesions of tuberculoid leprosy may sometimes make it
difficult to distinguish them from other infectious and noninfec-
tious granulomas, especially those of sarcoidosis and syphilis
[12]. Pt.R for the detection of M leprae in tissue lesions is
being developed, and this may in the future allow more accurate
diagnosis in such circumstances [13].
Buruli ulcer. M. ulcerans is the cause of chronic, relatively
painless, cutaneous Buruli ulcers. The disease is most prevalent
in Africa and Australia. The organism causes extensive under-
mined ulcers on the extensor surface of the extremities. The
centers of the ulcers are necrotic, and the edges are undermined;
the organisms are usually found at the periphery, where granu-
lation tissue is most extensive [14]. While it is relatively easy
to diagnose Buruli skin ulcers on the basis of clinical features
em 1996;23 (July)
Mycobacterial species
M. avium complex, M. kansasii, M. xenopi,
M. simiae, M scrofulaceum, M. chelonae,
M. malmoense
M. avium-M. intracellulare, M kansas ii,
M fortuitum, M. scrofulaceum,
M. chelonae
M. avium-M. intracellulare, M. kansasii,
M fortuitum, M scrofulaceum,
M. marinum
M chelonae, M. fortuitum
M. ulcerans
M. marinum
M. leprae
M. leprae
M. avium-M. intracellulare, M. kansasii,
M. chelonae
and histologic findings, microbiological identification of the
causal mycobacteria may sometimes be quite difficult, requir-
ing long periods of culture.
Newer techniques such as gas-phase chromatography are
becoming useful for identification of the acid-fast bacilli in
low-count subcultures [15]. A report from West Africa [16] of
a case of disseminated M. ulcerans infection that followed a
snakebite and caused multifocal osteomyelitis illustrates the
difficulties that are associated with identifying the organism
by conventional microbiological techniques. Although repeated
cultures were negative, the organism was identified as M. ul-
cerans by peR amplification of the genes coding for 168 rRNA
from biopsy specimens.
Swimming pool (fish tank) granuloma. M marinum causes
swimming pool (fish tank) granuloma [17]. Although the pri-
mary skin infection may be inconspicuous, the draining lymph
nodes are extensively involved and caseous. A similar micro-
scopic picture, with conspicuous plasma cell infiltration, is as-
sociated with granulomas due to other opportunistic mycobac-
teria. Fish tank granulomas develop in persons with minor
abrasions who dip their hands in tropical fish tanks. Usually a
solitary granuloma, nodule, or pustule forms, which may ulcer-
ate or suppurate; however, multiple lesions may extend along
the line of lymphatic vessels (in sporotrichotic infections).
Biopsy specimens that are cultured on Lowenstein-Jensen
medium at room temperature yield pigmented mycobacterial
colonies in 2~4 weeks. The response to treatment is variable
and not dramatic. Antituberculous drugs, cotrimoxazole, and
high doses of minocycline have been advocated.
A wide range of mycobacterial diseases occur in humans
(table 2), and in some cases the clinical features and results of
staining for acid-fast bacilli in granulomas may not reveal the
identity of the causal organisms, which may also be difficult
em 1996;23 (July) Granulomatous Infections
to isolate in culture. The development and application of molec-
ular techniques such as PCR may in the future allow fairly
accurate diagnosis [18].
Bacterial Infections
Brucellosis. Brucellosis (Malta fever) is a multisystem
granulomatous disorder caused by small gram-negative, intra-
cellular coccobacilli of the Brucella species [19]. The United
Kingdom is one of 17 Brucella-free countries. It is more preva-
lent than its reported incidence in the United States, France,
Spain, and Ireland [20]. Unpasteurized dairy products remain
a common mode of transmission of this chronic granulomatous
disorder among travelers to areas of endemicity. Raw camel's
and goat's milk, raw sheep's and goat's liver, and reindeer
bone marrow have all been associated with transmission.
Diagnosis, which may be difficult by means of culture, is
usually based on clinical features and serological and histopath-
ologic findings [20]. Delayed-type hypersensitivity is believed
to be important in the formation of the tissue granuloma, which
limits the spread of the organism [21]. Epithelioid-cell granulo-
mas may be found in the reticuloendothelial system, lung, brain,
bone, joints, and soft tissue and are difficult to distinguish from
granulomas due to other causes.
Necrotizing hepatic granulomas due to brucellosis have been
described. In these cases, bacteriologic diagnosis may be diffi-
cult on the basis of blood or abscess pus cultures, although
serology may be useful [22]. PCR tests for the identification
of Brucella species are being developed [23, 24], and these
may be useful tools for the diagnosis of brucellar granulomas.
Melioidosis. Melioidosis is caused by a gram-negative ba-
cillus, Burkholderia pseudomallei. Acute melioidosis presents
with protean clinical manifestations during which septicemia
may occur, causing septic shock and multiorgan involvement.
This may progress to a chronic suppurative phase, in which
lung involvement, subcutaneous abscesses, lymphadenitis, and
suppurative parotitis are common [25, 26]. Granulomatous os-
teomyelitis has also been described [27].
The lesions vary histologically from acute, chronic inflam-
mation with a focal granulomatous component to a predomi-
nantly granulomatous picture. Caseous or purulent central ne-
crosis is surrounded by epithelioid and giant cells and by
fibrosis in lungs, lymph nodes, and bone. The presence of
intracellular gram-negative bacteria within macrophages or gi-
ant cells and of similar extracellular bacilli may provide a clue
to the diagnosis [28].
More-specific enzyme immunoassays are being developed
for the detection of circulating antibody [29] or of B. pseu-
domallei antigen in the urine [30] of patients with melioidosis.
PCR tests for the identification of B. pseudomallei have re-
cently been described [31, 32], but these require field testing
in areas of endemicity.
149
Treponemal Infections
The spirochetes include Treponema pallidum subspecies pal-
lidum (the cause of syphilis), T. pallidum subspecies pertenue
(responsible for yaws), and Treponema carateum (the causal
agent of nonvenereal pinta in Latin America). All three are
chronic granulomatous disorders that may cause diagnostic
confusion with other granulomatous disorders [33]. Mucocuta-
neous lesions of secondary syphilis show a spectrum of
histopathologic changes, ranging from minimal infiltration to
granulomatous infiltration throughout the dermis [34]; granulo-
matous infiltrates show endothelial proliferation, with mononu-
clear cell infiltration.
The syphilitic gumma may present as a microscopic to
grossly visible lesion, with an enclosing wall of histiocytes,
plasma cell infiltrate, and necrotic center cells with intact cellu-
lar outlines. peR to detect T. pallidum subspecies pallidum
DNA in clinical specimens, including gumma biopsy material,
appears to be highly sensitive and specific [35, 36].
Protozoan Infections
Leishmaniasis. Leishmaniasis refers to the spectrum of
clinical disease caused by the protozoan Leishmania [37]. This
parasite is transmitted by phlebotomine sandflies, and the dis-
ease presents in three clinical forms: cutaneous, mucocutane-
ous, and visceral. Amastigotes of Leishmania species live
within macrophages and are usually seen as single or focal
collections within macrophages in blood, aspirates of sternal
marrow, liver or spleen specimens, or special culture medium.
Leishmaniasis has been extensively studied as a model infec-
tion for analysis of cell-mediated immunity. Granulomatous
responses are a feature of cutaneous and mucocutaneous leish-
maniasis. Cutaneous ulceration is characterized by a mononu-
clear cell infiltrate, with Th l-type responses emerging after
weeks of infection. Resolution of the infection depends on an
increase in the number of leishmania-specific CD4 + T cells of
the Thl subset [38], following which a granulomatous response
with epithelioid and giant cells emerges.
The nodular stage of post-kala-azar dermal leishmaniasis is
characterized by massive granulomas consisting of lympho-
cytes, plasma cells, and histiocytes, with numerous amastigotes
of the species Leishmania donovani [39]. A preponderance of
CD8+ T cells occurs in the lesion, in contrast to normal levels
in the peripheral blood. A recent report described the use of
PCR for identifying Leishmania aethiopica DNA in formalin-
fixed and paraffin-embedded tissue specimens [40]: seven of
the 40 skin biopsy specimens in which parasites were not found
histopathologically were subsequently found to contain para-
site DNA.
Toxoplasmosis. Toxoplasmosis refers to the clinical dis-
ease caused by the protozoan Toxoplasma gondii. Neonatal
infection occurs during acute maternal infection in pregnancy.
Acquired toxoplasmosis presents as a glandular, fever-like syn-
150 Zumla and James em 1996;23 (July)

Table 3. Causes and immunopathologic features of granulomatous mycoses.

Fungal species Clinical condition Immunopathologic feature(s)

Cryptococcus neoformans Cryptococcosis Pneumonia, infarction, abscess,

Candida species
Sporothrix schenckii
Histoplasma capsulatum
varieties
Aspergillus species
Paracoccidioides brasiliensis
Coccidioides immitis
Blastomyces dermatitidis
Phialophora species
Pseudallescheria boydii,
Madurella species
meningitis, granuloma, fibrosis
Candidiasis Abscess, necrosis, granulomas
(mucocutaneous or multisystem)
Sporotrichosis Granuloma (cutaneous, skeletal)
Histoplasmosis Pneumonia, cavitation, granuloma
Aspergillosis Necrotizing multisystem granulomas
South American blastomycosis Pneumonia, cavitation, granuloma
Coccidioidomycosis Pneumonia, cavitation, granuloma,
cutaneous plaques, nodules
Blastomycosis Granuloma, microabscess, pneumonia,
chronic pulmonary or extrapulmonary
disease
Chromoblastomycosis Granuloma (cutaneous)
Mycetoma Granuloma (skin and subcutaneous tissue)

drome that rapidly resolves. The T. gondii trophozoites encyst
and persist in nucleated cells as the cause of chronic, latent
infection that can reactivate with immunosuppression and pro-
duce multisystem disease.
The histopathologic changes characteristically occur with a
triad of features [41]: reactive follicular hyperplasia, epithelioid
histiocytes, and monocytoid cells. Langhans' giant cells are
not typically seen. Accurate diagnosis may be difficult in many
cases. PCR is increasingly being used to detect T. gondii DNA
in various clinical samples [42], and its application to identifi-
cation of T. gondii DNA in tissue biopsy specimens is under
investigation.
Fungal Infections
Coccidioides immitis spores inhaled by immunocompetent
persons induce a delayed-type hypersensitivity reaction. Most
primary cases of coccidioidomycosis involving such patients
are asymptomatic. Lung lesions may develop in some patients,
in association with fever, chest symptoms, and erythema nodo-
sum. A minority of cases progress to disseminated infection.
The primary and secondary granulomatous lung lesions due
to C. immitis are similar to those due to Histoplasma species.
C. immitis organisms are thick-walled, nonbudding spherules
often filled with endopores. A neutrophil infiltrate may occur
around the area of granulomatous reaction when the spherules
rupture to release the endospores. In disseminated disease, the
inflammatory response may be purely granulomatous, pyo-
genic, or mixed. Pyogenic lesions may dominate in immuno-
suppressed patients [46, 47].

Fungal infections are common causes of granulomas world-

wide. Granulomatous fungal infections can present as localized
conditions or systemic illnesses (table 3). A wide range of
medically important fungi can now be diagnosed by PCR [43].
Histoplasmosis and coccidioidomycosis are caused by dimor-
phic fungi that produce clinical disease resembling tuberculosis
[44-46].
Histoplasma infections in the lungs of immunocompetent
persons cause epithelioid-cell granulomas that undergo coagu-
lative necrosis. Histologic differentiation from tuberculosis,
sarcoidosis, and coccidioidomycosis may be possible by identi-
fication of thin-walled yeast forms in Grocott-Gomori methena-
mine-silver nitrate stains. In immunodeficient persons, histo-
plasmosis may be fulminant and is not associated with
formation of epithelioid-cell granulomas: focal collections of
mononuclear phagocytes containing yeasts are seen throughout
tissues of the body, and the reticuloendothelial system becomes
full of macrophages containing yeasts [46, 47].
Helminthic Infections
Several helminthic parasites can cause granulomas with
pathological consequences. Of particular interest are schistoso-
miasis and visceral larva migrans.
Schistosomiasis. Schistosomiasis is caused by five main
human-blood flukes of the genus Schistosoma: S. haematob-
ium, S. mansoni, S. japonicum, S. intercalatum, and S. mekongi.
These currently infect more than 200 million people. Diagnosis
is made by identification of the species-specific characteristic
ova in feces, urine, or tissues (usually rectal and liver biopsy
specimens). During the chronic phase of the infection, the ma-
ture worms produce large numbers of ova, which become
trapped in tissues and induce formation of granulomas [48],
with chronic fibro-obstructive sequelae.
Granuloma formation is the result of a combination of fac-
tors: a foreign-body reaction to the deposited ova (which may
em 1996;23 (July) Granulomatous Infections
or may not be identifiable within the granuloma) and a cell-
mediated delayed-type hypersensitivity reaction to antigenic
determinants of the parasite [49]. The giant cells are of the
multinucleate type, and the eosinophils are conspicuous. The
ova contain living larvae that release a variety of toxic and
antigenic substances through the egg wall, which itself is anti-
genic.
These substances induce and maintain the granulomatous
response. The granulomatous response appears to be tightly
regulated and involves development of Th l-type and Th2-type
cytokine responses [50, 51].
Toxocariasis (visceral larva migrans). Dog and cat larval
nematodes, especially Toxocara canis and Toxocara cati, may
infect children and lead to the syndrome of visceral larva mi-
grans [52]. Swallowed infected eggs emerge as larvae in the
intestine, whence they pass to the liver, lungs, brain, and eye,
producing eosinophilic granulomas in these different organs.
This infection should be particularly suspected in children with
hepatomegaly or miliary granulomas in liver biopsy specimens,
miliary lung mottling, eosinophilia, and focal posterior uveitis
or endophthalmitis. Liver granulomas may reveal part of the
larvae with surrounding eosinophilia.
Viral Infections
Measles. The paramyxovirus family is a group of RNA
viruses that includes the measles virus, which has been impli-
cated in the etiology of idiopathic granulomatous disorders,
sarcoidosis, and Crohn's disease. The measles (rubeola) virus
is spread by respiratory droplets and it multiplies within the
upper respiratory tract epithelial cells, mononuclear cells, B
and T lymphocytes, and macrophages. Viremia follows, which
disseminates the virus throughout the body. T cell-mediated
immunity that controls the infection develops and produces the
measles rash.
The blotchy rash of measles consists of dilated skin vessels,
edema, and a nonspecific mononuclear perivascular infiltrate
[47,53]. Lymphoid organs have marked follicular hyperplasia,
large germinal centers, and randomly distributed multinucleate
giant cells (Warthin-Finkeldey cells), which have eosinophilic
nuclear and cytoplasmic inclusion bodies. In the lung, peribron-
chial and interstitial mononuclear infiltration occurs. The mea-
sles virus has been implicated in the etiopathogenesis of sar-
coidosis and Crohn's disease and is discussed in the section
on group 3 disorders.
Herpesviruses. The herpes group of viruses are double-
stranded DNA viruses that cause a spectrum of human disease.
The Epstein-Barr virus is the etiologic agent of the clinical
condition infectious mononucleosis and is transmitted by close
contact, especially via saliva during kissing. It has also been
implicated in the pathogenesis of several disorders such as
Burkitt's lymphoma, nasopharyngeal carcinoma, B cell
lymphomas, and sarcoidosis.
151
Infectious mononucleosis in its benign form is characterized
by fever, sore throat, lymphadenopathy, splenomegaly, and the
appearance of atypical activated T cells in the peripheral blood.
Further multisystem involvement may occur, leading to hepati-
tis, meningoencephalitis, and pneumonitis. Atypical lympho-
cytes are seen in the portal areas and sinusoids of liver biopsy
specimens, and scattered isolated areas of focal parenchymal
necrosis filled with lymphocytes may occur [47].
This histologic picture is difficult to distinguish from that of
other types of viral hepatitis, such as hepatitis B, and from that
of cytomegalovirus [54] infections. Specific diagnosis can be
made by culture of the virus or identification of virus-specific
DNA by PCR [55].
Group 2 Granulomatous Disorders (Recently Recognized
Causal Agents)
Cat-Scratch Disease
Cat-scratch disease or fever is also known as benign lymph-
oreticulosis or regional granulomatous lymphadenitis [56]. It
only occurs in humans, especially those who are scratched
or bitten by kittens and then develop regional lymphadenitis
proximal to the site of injury. Primary involvement is that of
the lymph nodes, which first show lymphoid hyperplasia. Later,
scattered granulomas with central areas of necrosis coalesce to
form abscesses.
The histopathologic features of cat-scratch disease are not
diagnostic and may be mistaken for tularemia, lymphogranu-
loma venereum, syphilis, brucellosis, atypical mycobacterial
infections, fungal infections, and toxoplasmosis [47]. Warthin-
Starry silver staining is used to detect Bartonella henselae,
which may be present in the early phase of the disease. A skin
test antigen has been made from lymph node pus. It is inocu-
lated intradermally, and the degree of induration and erythema
is measured at 48 hours.
The cat-scratch antigen skin test is positive for about 90%
of patients who are clinically suspected of having the disease.
This test will become redundant when techniques for ampli-
fying specific nucleotide sequences with peR come into gen-
eral use. There is no well-recognized response to antibiotics,
and recovery usually occurs without treatment.
Until recently, the causal agent remained controversial. Two
organisms, Afipia felis and B. henselae, have been identified
as potential candidates [57]. Substantial evidence is now accu-
mulating that B. henselae, a small pleomorphic gram-negative
bacillus, is the etiologic agent [58-60].
A PCR hybridization assay designed to amplify DNA from
B. henselae and A. felis was used on pus aspirates from 89
skin test-positive patients with cat-scratch disease. B. henselae
DNA was found in 96% of samples, while no samples con-
tainedA.felis DNA [58]. An indirect fluorescent antibody assay
for B. henselae-specific antibody has been described that
appears to be sensitive for the diagnosis of cat-scratch dis-
ease [61].
152 Zumla and James em 1996;23 (July)

Whipple's Disease graft-versus-host rejection, with similar structural changes in

the bile, lacrimal, and pancreatic ducts, which have a high
George Hoyt Whipple of Johns Hopkins Hospital initially
concentration of human leukocyte class II antigens on the epi-
described a 37-year-old missionary who presented with fever,
thelial surface.
polyarthritis, and steatorrhea. His report was entitled "A Hith-
PBC is distinguished by the fact that it predominates in
erto Undescribed Disease Characterized Anatomically by De-
women in the reproductive years of age and by the presence
posits of Fat and Fatty Acids in the Intestinal and Mesenteric
of serum mitochondrial antibodies. It is classified as an autoim-
Lymphatic Tissues" [62]. Whipple's disease is a chronic
mune disorder and is associated with other autoimmune condi-
multisystem granulomatous disorder that affects middle-aged
tions, including rheumatoid arthritis, systemic lupus erythema-
white males. It presents, as did Whipple's patient, with fever,
tosus, scleroderma, Sjogren's syndrome, and the CREST
polyarthritis, weight loss, and diarrhea progressing to malab-
syndrome.
sorption.
The etiology of PBC remains unknown. Two infectious
Hepatosplenomegaly and generalized lymphadenopathy may
agents have been postulated as possible trigger cofactors [74,
be present, and biopsy of the lymph node, liver, or small intes-
75]. The antigen specific for PBC serum is M2, a component
tine will reveal sarcoid granulomas, of which PAS stains will
of the pyruvate dehydrogenase complex of mitochondrial en-
reveal foamy macrophages. The PAS-positive material corre-
zymes. There are similarities between mitochondrial compo-
sponds with lysosomes containing bacilliform bodies. Electron
nents and Escherichia coli, and cross-reactivity between bile
microscopy reveals rod-shaped bacilli, termed Whipple bacilli
duct mitochondria and bacteria may be conceivable [73]. Mito-
[63], Whipple's associated bacterial organisms, or (more cor-
chondrial antibodies cross-react with subcellular constituents
rectly) Tropheryma whippelii [64-66].
or gram-negative intestinal or urinary tract organisms.
The nucleic acids extracted from an endoscopic biopsy speci-
An increased incidence of gram-negative urinary tract infec-
men of the proximal small bowel of a patient with Whipple's
tions in patients with PBC has been reported. Another infective
disease were subjected to amplification and nucleotide sequenc-
association was suggested in a report from Barcelona that in-
ing by PCR. The resulting PCR product from the bacterial 16S
criminated Mycobacterium gordonae. This ubiquitous organ-
rRNA was then the subject of a computer database search for
ism causes granuloma formation, and serum antibodies are evi-
the rRNA sequences most similar to it.
dent in PBC serum. Furthermore, there is cross-reactivity
T. whippelii is an actinomycete that behaves like an intracel-
between this organism and the important PBC-defining M2
lular pathogen and responds particularly well to antibiotics that
antigen. PCR has disclosed the presence of this mycobacterium
enter cells easily. It is an erythrocyte-associated organism (like
in the tissues and bile of patients with PBC [75].
Bartonella bacilliformis, Babesia species, and Plasmodium
species, which are also hemotropic). T. whippelii DNA has
been detected by PCR examination of peripheral blood, pleural Sarcoidosis
effusion cells [66, 67], biopsy tissue from the intestine [68],
Sarcoidosis is a multisystem disorder of unknown etiology
heart [69], or vitrectomy specimen of the eye [70].
[76]. An antigen (chemical or infectious organism)-driven,
The PCR primers that have been developed are specific for
cell-mediated immune response leads to a cytokine cascade, to
T. whippelii, and this technique is now one of the standard
sarcoid granuloma formation, and eventually to sarcoid fibrosis.
methods for establishing the diagnosis of Whipple's disease
An intense search for an infectious cause has yielded many
[66, 67, 70]. The technique ensures a more specific diagnosis
theories, yet none is conclusive (table 4).
since the organism cannot be cultivated. Other data show that
In 1961, Edith Mankiewicz [77], of Montreal, reported that
there may be a closely related second causative agent of Whip-
bacteriophages, lytic for mycobacteria, can be isolated with
ple's disease [71, 72].
higher frequency from stool and resection specimens from pa-
tients with tuberculosis and sarcoidosis. Raised titers of phage-
Group 3 Granulomatous Disorders (Infective Agent neutralizing antibodies were found in tuberculous patients in-
Strongly Suspected) fected with mycobacteria harboring mycobacteriophages but
not in cases of sarcoidosis.
Primary Biliary Cirrhosis
While mycobacteriophages are common to both diseases
This condition of progressive, nonsuppurative, granuloma- the persistence of phages without antibody was the explanation
tous destruction of intrahepatic bile ducts has also been termed given for the absence of M. tuberculosis in sarcoid tissue [78].
xanthomatous biliary cirrhosis, but its most accurate description This observation led to further experiments in which guinea
is chronic nonsuppurative destructive cholangitis [73]. Liver pigs were infected with tubercle bacilli alone or together with
cell necrosis may become superadded, later progressing to cir- mycobacteriophages, and the histological responses (analyzed
rhosis. Epithelioid granulomas associated with the bile ducts blindly) were of two dissimilar types. The first was caseous
are found in about one-third of the cases of primary biliary necrosis with the presence of acid-fast bacilli, and the second
cirrhosis (PBC). The condition has been likened to chronic was a different response modified by the mycobacteriophages:
cm 1996; 23 (July) Granulomatous Infections 153

Table 4. Causal agents implicated in cases of sarcoidosis. The number of hybrids obtained with the sarcoid spleens (from
which mycobacteria were neither seen on microscopy nor cul-
Causal agents implicated Reference( s)
tured) was 4.8 times higher than that obtained with normal
Bacteria spleens.
Mycobacteria [76-92] However, the results obtained by PCR are being fiercely
Streptococci [94] contested by German [83], Scottish [84], and French [85] inves-
Propionibacterium acnes [88,93] tigators. The findings of the German group contradict those of
Borrelia burgdorferi [95,96]
Saboor et a1. [81], and the group concludes that mycobacterial
Mycoplasma species [97]
Nocardia species [98] DNA cannot be detected in tissues or bronchoalveolar lavage
Viruses cells in cases of sarcoidosis. They use the technique to distin-
Epstein-Barr [99] guish tuberculosis (which produces a positive result) from sar-
Herpesvirus [100, 101] coidosis. The Glasgow investigators [84] also failed to detect
Rubella [102]
mycobacterial DNA in sarcoid lymph nodes. The French inves-
Measles [102]
Cytomegalovirus [102] tigators [85], using PCR, rarely found DNA from M. tuberculo-
Coxsackievirus B [102,103] sis in cases of sarcoidosis.
Retrovirus [104] The literature for and against a mycobacterial cause of sar-
Others coidosis is summarized in table 5. The existence ofmycobacte-
Beryllium [lOS]
rial sarcoidosis would validate several claims of infective or
Zirconium [105]
Pine pollen [106,107] chemical causes of sarcoidosis: Scadding's long-held views on
Peanut dust [107] tuberculous causation [86], a Swedish suggestion [87] of an
Clay (ingestion) [108] interaction between mycobacteria and a virus, a Japanese the-
ory [88] regarding Propionibacterium as the causal agent, and
that of investigators in Houston [89] who isolated cell wall-
the phage-infected animals showed a granulomatous sarcoid- defective spheroplasts of acid-fast mycobacteria from sarcoid
like picture with very few atypical lysogenic mycobacteria. skin biopsy specimens.
The conclusion reached was that phages of different anti- The controversy continues [90-93] and many different anti-
genic composition determined a shift in the antigen-antibody gens have come under suspicion, including other bacteria [94-
response toward either caseation or noncaseating granuloma 97], fungi [98], viruses [99-104], and chemicals [105-- 108].
formation. It seemed at that time that sarcoidosis resulted from Many attempts have been made but tissue culture has failed to
unrestrained lytic phage activity against mycobacteria [77, 78]. uncover a virus. Elevated antibody responses to several viruses
At about the same time, Chapman and Speight [79] reported have been reported, but such responses may only reflect B-cell
a high incidence of serum antibodies to mycobacteria in pa- hyperreactivity rather than a causal relationship.
tients with sarcoidosis. Serum antibody levels were particularly
high against photochromogen and scotochromogen antigens.
Crohn's Disease
The patients' antibody status changed from negative to positive
as sarcoidosis manifested itself, and in many cases antibodies Crohn's disease is a chronic granulomatous disorder affect-
appeared 2 months following the onset of erythema nodosum. ing predominantly the gut, but it has multisystem involvement
Ghosts of mycobacteria have been reported to be detected and variable clinical manifestations. The pathological changes
by fluorescent microscopy of scalene lymph node biopsy speci- may involve any parts and any layer of the alimentary tract.
mens from patients with sarcoidosis and from controls. Aur- Transmural inflammation consists of chronic inflammatory
amine-rhodamine stains disclosed shining golden rods resem- cells with lymphoid aggregates scattered throughout. Noncase-
bling mycobacteria in specimens from patients with sarcoidosis ating sarcoid-like granulomas may be present in all layers of
but not in those from controls [80]. Recent advances in molecu- diseased and unaffected tissue throughout the alimentary
lar biology have resurrected these observations; PCR has de- tract [109].
tected mycobacterial DNA in clinical samples from patients As with sarcoidosis, controversy prevails concerning a causal
with sarcoidosis [81, 82]. agent. Clinicopathologic studies at the Royal Free Hospital
Bronchoalveolar lavage samples, bronchial washings, and (London) demonstrated the presence of granulomatous vasculi-
tissue specimens have been assayed by PCR for mycobacterial tis and vessel thrombosis, giving rise to focal microulcers. The
DNA. M. tuberculosis DNA has been found in 50% of patients investigations have shown early superficial mucosal changes
with sarcoidosis, and nontuberculous mycobacterial DNA has and disruption of the capillary basement membrane preceding
been found in 70% of them. Another study extracted RNA the formation of the microulcer [I 10]. These inflammatory
from five sarcoid and five normal spleens [81]. Extracts were changes may extend beyond the alimentary tract to the lung,
assayed by liquid-phase DNA/RNA hybridization with a DNA causing pulmonary vasculitis, granulomatous interstitial lym-
probe specific for the rRNA of the M. tuberculosis complex. phocyte infiltration, and interstitial fibrosis [II I, 112].
154 Zumla and James ern 1996;23 (July)

Table 5. Summary of the literature for and against a mycobacterial cause of sarcoidosis.

Laboratory method used or feature

Reference( s) Year(s) of study Location of study analyzed (finding)

[77,78] 1961-1964 Montreal Reaction to mycobacteriophage infection

[79] 1964 Dallas Presence of mycobacterial antibodies in
serum (positive)
[80] 1971 Prague Auramine-rhodamine staining,
fluorescent microscopy (positive)
[87] 1972 Stockholm Mycobacteria-virus interaction
[93] 1984 Tokyo Culture of lymph nodes (positive for
Propionibacterium acnes)
[89] 1988 Houston Skin biopsy (cell-wall-defective, acid-
fast bacilli noted)
[81] 1992 London PCR (positive for mycobacterial DNA)
[82] 1992 London Liquid-phase hybridization (positive for
mycobacterial rRNA)
[83] 1992 Borstel, Germany PCR (negative for mycobacterial DNA)
[84] 1992 Glasgow PCR (negative for mycobacterial DNA)
[85] 1992 Paris PCR (negative for mycobacterial DNA)
[90] 1993 London PCR (positive for mycobacterial DNA)
[91] 1993 Ohio ELISA (positive for antibodies to
Mycobacterium paratuberculosis)
[92] 1993 Copenhagen Western blotting (positive for antibodies
to Mycobacterium tuberculosis)

Against this background is the conflict about whether the
causal agent is Mycobacterium paratuberculosis [113], other
mycobacteria [114], or the measles virus [115, 116]. Measles
virus nucleocapsids have been detected in foci of granuloma-
tous inflammation of the intestine in cases of Crohn's disease
[117]. It has been postulated that Crohn' s disease may be a
result of chronic granulomatous vasculitis that develops in reac-
tion to a persistent infection with measles virus within the
vascular endothelium [118].
Langerhans' Cell Granulomatosis
The term Langerhans' cell granulomatosis refers to prolifera-
tive disorders of histiocytes, previously referred to as histio-
cytosis X [119]. It encompasses a group of disorders of un-
known etiology characterized by granulomatous infiltration of
the lungs, bone, skin, lymph nodes, and brain. The clinical
conditions have been known by several names, based on the
type of presentation, sites of involvement, rate of progression,
and degree of associated immune dysfunction. They include
eosinophilic granuloma, Letterer-Siwe disease, and Hand-
Schuller-Christian disease. These are now considered to be
different expressions for the same basic disorder, in which the
proliferation of Langerhans' cells results from disturbances in
immunoregu1ation.
Langerhans' histiocytes are bone marrow-derived mono-
cyte-macrophage cells; they include Langerhans' epidermal
cells, Kupffer's cells in the liver, osteoclasts, and alveolar mac-
rophages. They are human leukocyte antigen-DR-positive
functioning macrophages that present antigen to T cells and
play a role in cell-mediated immunity. Unlike histiocytes,
Langerhans' cells can be stained immunohistochemically for
S-100 protein and OKT-6.
Lung biopsy reveals a mixed celullar exudate, foam cells,
eosinophils, and characteristic X bodies (Birbeck granules) in
macrophages. Langerhans' or X bodies are an ultrastructural
feature in 90% of patients [120]. They are identical to the
granules in Langerhans' epidermal cells and consist of intracy-
toplasmic rod-, plate-, or cup-like pentalaminar structures.
The presence of these tennis racket-shaped ultrastructural
Birbeck granules is diagnostic of the disorder. They have sur-
face adenosine triphosphate activity identifiable by gold fluo-
rescence. These diagnostic cells are readily found by bron-
choalveolar lavage, and this technique may make lung biopsy
unnecessary. It may also be a likely means of detecting a
possible causal agent.
Chronic Granulomatous Disease of Childhood (CGD)
CGD comprises a genetically heterogeneous group of dis-
eases that have in common the functional disorder of a multi-
component enzyme system, NADPH oxidase, found in phago-
cytic cells [121-124]. This defect leads to severe infections,
particularly with Staphyloccocus aureus, Serratia marcescens,
Burkholderia cepacia, and Aspergillus species. Organisms that
lack catalase supply the neutrophil with the hydrogen peroxide
for their own destruction. Thus, catalase-negative organisms
such as pneumococci or streptococci present no problem be-
cause they are normally destroyed.
em 1996;23 (July) Granulomatous Infections
There are three autosomal recessive types, corresponding to
mutations in p22p h0X, p47p h0X, and p67p hox . The classic X-linked
disorder (found in ,....,60% of cases), which affects the gene
encoding gp91phox, occurs in boys and is diagnosed in infancy,
usually on the basis of an infection with a catalase-producing
bacterium or fungus. The diagnosis is confirmed on the basis
of a nitroblue tetrazolium test or other test showing defective
superoxide production.
Hepatosplenomegaly is common and does not reflect active
infection per se. Lymphadenopathy is often undetected, but
when it is significant it usually reflects an active infection.
Severe granulomatous problems can occur in association with
CGD, such as obstruction of the gut and the genitourinary
system. Weeping granulomatous skin lesions occur in the set-
ting of an active wound, typically post-surgically. The histolog-
ical appearance is marked by lipid-laden macrophages. While
the etiology of CGD is genetic, there may be undetected patho-
gens responsible for some of the granulomatous complications
of the disease.
Orofacial Granulomatosis (Melkersson-Rosenthal Syndrome)
This is a rare granulomatous disorder of the mouth and
adjacent tissues, involving the oral mucosa, gum, lips, tongue,
pharynx, eyelids, and skin of the face. Melkersson [125] de-
scribed an association between facial edema and facial paraly-
sis. Rosenthal [126] added the features of lingua plicata or
scrotal tongue. Other clinical features include granulomatous
cheilitis, edema of the gums and scalp, salivary gland dysfunc-
tion, granulomatous blepharitis, trigeminal neuralgia, Ray-
naud's phenomenon, and chronic hypertrophic granulomatous
vulvitis [127-129].
Biopsy of an involved area reveals a noncaseating granu-
loma. Unlike the findings in cases of sarcoidosis, there are no
chest radiographic changes or uveitis, and the Kveim-Siltzbach
skin test is negative. The etiology remains unknown.
Kikuchi's Disease
This disorder was described in 1972 by a Japanese patholo-
gist and is characterized by lymphadenitis with focal reticulum-
cell hyperplasia, nuclear debris, and phagocytosis [130]. The
lymphadenitis may be proliferative, necrotizing, or xanthoma-
tous. Immunohistologic analysis shows that the predominant
cells involved in the lymphadenitis are various types of histio-
cytes, plasmacytoid monocytes, and T cells; B cells are absent
[131]. Clinically there is localized, tender, cervicallymphade-
nopathy with an upper respiratory tract prodrome.
Kikuchi's disease was initially thought to occur mostly in
women under the age of 30 years, although a recent study of
79 cases showed that the preponderance of cases involving
females is not striking [131]. The disease runs a self-limiting
course and is associated with a recurrence rate of3%. Kikuchi's
disease has been recognized as a presenting feature of mixed
155
connective-tissue disease and has been noted in association
with adult Still's disease and systemic lupus erythematosus, a
finding leading to the hypothesis that Kikuchi's disease and
autoimmune rheumatic disorders may share a common etiol-
ogy [132].
The disease occurs worldwide and has often been confused
with toxoplasmosis, cat-scratch disease, tuberculosis, infectious
mononucleosis, and non-Hodgkin's lymphoma. The cause of
the disease is not known. A viral etiology is strongly suspected
on the basis of clinical features, although serological and
ultrastructural studies have not yet identified an infectious
agent [133].
Conclusions
Classification schemes are most useful when they provide
insight into the purpose of events and the etiologic mechanisms
that govern them. Granulomas are remarkably complex in-
flammatory foci that sequester organisms or other substances
resistant to degradation. While a large number of granuloma-
tous disorders are recognized, infections are clearly the most
common underlying causes of granulomas in general.
It is apparent that granulomatous conditions ofdiverse etiolo-
gies share common histologic features, although the etiologic
agent is not always identifiable. Although the histopathologic
patterns in various infectious granulomas may be sufficiently
different to prevent an accurate diagnosis, atypical presenta-
tions may necessitate identification of the specific etiologic
agent by direct microscopic examination, culture, serology, or
molecular detection.
In several granulomatous diseases the etiologic agent is dif-
ficult to identify by microscopic examination, culture, or sero-
logical means. The availability of PCR for the diagnosis of
several infectious conditions has been of major importance in
detecting the etiologic agents of granulomatous disease condi-
tions that previously were considered to be of unknown etiol-
ogy. In particular, PCR has been used successfully for detecting
the bacteriologic causes of Whipple's disease and cat-scratch
disease.
Further applications of molecular techniques may pin down
the microbiological etiology (if any) of several granulomatous
disorders of group 3 (table I), of which the etiology remains
elusive.
References
1. Warren KS. Granulomatous inflammation. In: Lepow IH, Ward PW, eds.
Inflammation: mechanisms and control. New York: Academic Press,
1972:203 -18.
2. Williams GT, Williams WJ. Granulomatous inflammation-a review. J
Clin Pathol 1983; 36:723 - 33.
3. Silverman L, Shorter RG. Histogenesis of the multinucleated giant cell.
Lab Invest 1963; 12:985-90.
4. James DG. What makes granulomas tick'? Thorax 1991;46:734-6.
156 Zumla and James
5. Heinzel FP. Th I and Th2 cells in the cure and pathogenesis of infectious
diseases. Curr Opin Infect Dis 1995; 8:151-5.
6. Weinstock JV. Function and regulation of granulomatous responses (ex-
perimental granulomatosis). In: James DG, ed. Sarcoidosis and other
granulomatous disorders. New York: Marcel-Dekker, 1994:99-133.
7. Yamamura M, Uyemura K, Deans RJ, et al. Defining protective responses
to pathogens: cytokine profiles in leprosy lesions. Science 1991;254:
277-9.
8. Cucin RL, Coleman M, Eckardt JJ, Silver RT. The diagnosis of miliary
tuberculosis: utility of peripheral blood abnormalities, bone marrow
and liver needle biopsy. J Chronic Dis 1973;26:355-61.
9. Schluger NW, Kinney D, Harkin TJ, Rom WN. Clinical utility of the
polymerase chain reaction in the diagnosis of infections due to Myco-
bacterium tuberculosis. Chest 1994; 105:1116-21.
10. Zumla A, James DG. Lung immunology in the tropics. In: Sharma OP,
ed. Lung disease in the tropics. New York: Marcel-Dekker, 1991:1-
64.
11. Lucas S, Nelson AM. Pathogenesis of tuberculosis in human immunode-
ficiency virus-infected people. In: Bloom BR, ed. Tuberculosis: patho-
genesis, protection and control. Washington, DC: American Society
for Microbiology, 1994:503-16.
12. Zumla A. Sarcoidosis and leprosy: a comparative perspective. In: James
DG, ed. Sarcoidosis and other granulomatous disorders. New York:
Marcel-Dekker, 1994:681-705.
13. Santos AR, Goes Filho JT, Nery JA, et al. Evaluation of PCR mediated
DNA amplification in non-invasive biological specimens for subclini-
cal detection of Mycobacterium leprae. FEMS Immunol Med Micro-
biol 1995;11:113-20.
14. Dodge OG. Mycobacterial skin ulcers in Uganda: histopathological and
experimental aspects. J Pathol Bacteriol 1964; 88:167- 74.
15. Delaporte E, Savage C, Alfandari S, Piette F, Leclerc H, Bergoend H.
Ulcere de Buruli chez une malade Zairoise infectee par le virus de
l'immunodeficience humaine. Ann Dermatol Venereol 1994; 121:
557-60.
16. Hofer M, Hirschel B, Kirschner P, et al. Brief report: disseminated osteo-
myelitis from Mycobacterium ulcerans after a snakebite. N Engl J
Med 1993; 328:1007-9.
17. Gray SF, Smith RS, Reynolds NJ, Williams EW. Fish tank granuloma.
Brit Med J 1990;300: 1069-1070.
18. Shinnick TM, Jonas V. Molecular approaches to the diagnosis oftubercu-
losis. In: Bloom BR, ed. Tuberculosis: pathogenesis, protection and
control. Washington, DC: American Society for Microbiology, 1994:
511-31.
19. Bruce D. Malta fever. BMJ 1889; 1:1101-7.
20. Trujillo IZ, Zavala AN, Caceres JG, Miranda CQ. Brucellosis. Infect Dis
Clin North Am 1994;8:225-41.
21. Adams DO. The biology of the granuloma. In: Ioachim HL, ed. Pathology
of granulomas. New York: Raven Press, 1983:1-20.
22. Debat-Zoguereh D, Badiaga S, Uzan E, Le Treut YP, Lebreuil G, Bour-
geade A. Granulome necrosant hepatique d'origine brucellienne. A
propos d'un cas. Rev Med Interne 1995; 16:63-6.
23. Romero C, Gamazo C, Pardo M, Lopez-Goff I. Specific detection of
Brucella DNA by PCR. J Clin Microbiol 1995; 33:615- 7.
24. Bricker BJ, Halling SM. Differentiation of Brucella abortus bv. 1, 2, and
4, Brucella melitensis, Brucella ovis, and Brucella suis bv. 1 by PCR.
J Clin Microbiol 1994;32:2660-6.
25. Leelarasamee A, Bovomkitti S. Melioidosis: review and update. Rev
Infect Dis 1989; 11:413- 25.
26. Lumbiganon P, Viengnondha S. Clinical manifestations of melioidosis
in children. Pediatr Infect Dis J 1995; 14:136-40.
27. SirikuIchayanonta V, Subhadrabandhu 1. Melioidosis: another etiology
of granulomatous osteomyelitis. Report of 2 cases. Clin Orthop 1994;
308:183-6.
28. Wong KT, Puthucheary SD, Vadivelu J. The histopathology of human
melioidosis. Histopathology 1995; 26:51- 5.
ern 1996;23 (July)
29. Rugdech P, Anuntagool N, Sirisinha S. Monoclonal antibodies to Pseu-
domonas pseudomallei and their potential for diagnosis of melioidosis.
Am J Trop Med Hyg 1995;52:231-5.
30. Desakom V, Smith MD, Wuthiekanun V, et al. Detection of Pseudomo-
nas pseudomallei antigen in urine for the diagnosis of meiloidosis.
Am J Trop Med Hyg 1994;51:627-33.
31. Kunakom M, Markham RB. Clinically practical seminested PCR for
Burkholderia pseudo mallei quantitated by enzyme immunoassay with
and without solution hybridization. J Clin Microbiol 1995;33:
2131-5.
32. Lew AE, Desmarchelier PM. Detection of Pseudomonas pseudomallei
by PCR and hybridization. J Clin MicrobioI1994;32:1326-32.
33. Tramont EC. Treponema pallidium (syphilis). In: Mandell GL, Bennett
JE, Dolin R, eds. Mandell, Douglas and Bennett's principles and prac-
tice of infectious diseases. 4th ed. Vol. 2. New York: Churchill Living-
stone, 1995:2117-33.
34. Pandhi RK, Singh N, Ramam M. Secondary syphilis: a clinicopathologic
study. Int J Dermatol 1995;34:240-3.
35. Burstain JM, Grimprel E, Lukehart SA, Norgard MV, Radolf JD. Sensi-
tive detection of Treponema pallidum by using the polymerase chain
reaction. J Clin Microbiol 1991;29:62-9.
36. Horowitz HW, Valsamis MP, Wicher V, et al. Brief report: cerebral
syphilitic gumma confirmed by the polymerase chain reaction in a
man with human immunodeficiency virus infection. N Engl J Med
1994;331: 1488-91.
37. Peters W, Killick-Kendrick R, eds. The leishmaniases in biology and
medicine. London: Academic Press, 1987.
38. Locksley RM, Louis JA. Immunology ofleishmaniasis. Curr Opin Immu-
nol 1992;4:413-8.
39. Ghosh MK, Nandy A, Addy M, Maitra TK, Ghose AC. Subpopulations
of T lymphocytes in the peripheral blood, dermal lesions and lymph
nodes of post kala-azar dermal leishmaniasis patients. Scand J Immu-
noI1995;41:1l-7.
40. Laskay T, Miko TL, Negesse Y, Solbach W, Rollinghoff M, Frommel
D. Detection of cutaneous Leishmania infection in paraffin-embedded
skin biopsies using the polymerase chain reaction. Trans Roy Soc Trop
Med Hyg 1995;89:273-5.
41. Dorfman RF, Remington JS. Value oflymph-node biopsy in the diagnosis
of acute acquired toxoplasmosis. N Engl J Med 1973;289:878-81.
42. Novati R, Castagna A, Morsica G, et al. Polymerase chain reaction for
Toxoplasma gondii DNA in the cerebrospinal fluid of AIDS patients
with focal brain lesions. AIDS 1994;8:1691-4.
43. Makimura K, Murayama SY, Yamaguchi H. Detection of a wide range
of medically important fungi by the polymerase chain reaction. J Med
Microbiol 1994;40:358-64.
44. Davies SF. Fungal pneumonia. Med Clin North Am 1994;78:1049-65.
45. Wong JS, Herman SJ, de Hoyos A, Weisbrod GL. Pulmonary manifesta-
tions of coccidioidomycosis. Can Assoc Radiol J 1994;45:87-92.
46. Wheat J. Histoplasmosis and coccidioidomycosis in individuals with
AIDS: a clinical review. Infect Dis Clin North Am 1994; 8:467-82.
47. Samuelson J, Von Lichtenberg F. Infectious diseases. In: Cotran RS,
Kumar V, Robbins SL, eds. Robbins pathologic basis of disease. Phila-
delphia: WB Saunders, 1994:305-77.
48. Warren KS. The pathology, pathobiology and pathogenesis of schistoso-
miasis. Nature 1978;273:609-12.
49. Warren KS, Domingo EO, Cowan RBT. Granuloma formation around
schistosome eggs as a manifestation of delayed hypersensitivity. Am
J PathoI1967;51:735-56.
50. Chikunguwo SM, Quinn JJ, Ham DA, Stadecker MJ. The cell-mediated
response to schistosomal antigens at the clonal level. III. Identification
of soluble egg antigens recognized by cloned specific granulomagenic
murine CD4+ Thl-type lymphocytes. J Immunoll993; 150:1413-21.
51. Wynn TA, Cheever AW, Jankovic D, et al. An IL-12 based vaccination
method for preventing fibrosis induced by schistosome infection. Na-
ture 1995;376:594-6.
em 1996;23 (July) Granulomatous Infections
52. Huntley CC, Costas MC, Lyerly A. Visceral larva migrans syndrome:
clinical characteristics and immunologic studies in 51 patients. Pediat-
rics 1965; 36:523-36.
53. Halsey NA, Job JS. Measles. In: Warren KS, Mahmoud AAF, eds. Tropi-
cal geographic medicine. New York: McGraw-Hill, 1994.
54. Heyries L, Perreard M, Monges 0, Chrestian MA, Gerolami A. Hepatic
granulomatosis associated with mononucleosis syndrome secondary to
cytomegalovirus infection: apropos of 2 cases in healthy adults. Rev
Med Interne 1994; 15:744-6.
55. Nahass GT, Penneys NS, Leonardi CL. Analysis of the polymerase chain
reaction in the detection of herpesvirus DNA from fixed and stained
tissue sections. Arch Dermatol 1995; 131:805-8.
56. Daniels WB, MacMurray FG. Cat scratch disease: report of one hundred
sixty cases. JAMA 1954; 154:1247-51.
57. Drancourt M, Raoult D. La maladie des griffes du chat et la pathologie
a Bartonella (Rochalimaea). Presse Med 1995;24:183-8.
58. Bergmans AMC, Groothedde J-W, Schellekens JFP, van Embden IDA,
Ossewaarde JM, Schouls LM. Etiology of cat scratch disease: compari-
son of polymerase chain reaction detection of Bartonella (formerly
Rochalimaea) and Afipia felis DNA with serology and skin tests. J
Infect Dis 1995; 171:916-23.
59. Anderson B, Sims K, Regnery T, et a1. Detection of Rochalimaea
henselae DNA in specimens from cat scratch disease patients by PCR.
J Clin Microbiol 1994; 32:942-8.
60. Goral S, Anderson B, Hager C, Edwards KM. Detection of Rochalimaea
henselae DNA by polymerase chain reaction from suppurative nodes
of children with cat-scratch disease. Pediatr Infect Dis J 1994; 13:
994-7.
61. Dalton MJ, Robinson LE, Cooper J, Regnery RL, Olson JG, Childs JE.
Use of Bartonella antigens for serologic diagnosis of cat-scratch dis-
ease at a national referral center. Arch Intern Med 1995; 155:1670-6.
62. Whipple GH. A hitherto undescribed disease characterized anatomically
by deposits of fat and fatty acids in the intestinal and mesenteric
lymphatic tissues. Johns Hopkins Hospital Bulletin 1907; 18:382-91.
63. Spapen HOM, Segers 0, De Wit N. Electron microscopic detection of
Whipple's bacillus in sarcoidlike periodic acid-Schiff-negative granu-
lomas. Digest Dis Sci 1989;34:640-3.
64. Reiman DA, Schmidt TM, Mac Dermott RP, Falkow S. Identification of
the uncultured bacillus of Whipple's disease. N Engl J Med 1992;327:
293-301.
65. Donaldson RM Jr. Whipple's disease-s-rare malady with uncommon
potential [editorial]. N Engl J Med 1992;327:346-8.
66. Dobbins WO III. The diagnosis of Whipple's disease [editorial]. N Engl
J Med 1995;332:390-2.
67. Lowsky R, Archer GL, Gyles G, et al. Brief report: diagnosis of Whip-
ple's disease by molecular analysis of peripheral blood. New Engl J
Med 1994;331:1343-6.
68. Maiwald M, Meier-Willersen HJ, Hartmann M, von Herbay A. Detection
of Tropheryma whippelii DNA in a patient with AIDS. J Clin Microbiol
1995;33:1354-6.
69. Wendler 0, Mendoza E, Schleiffer T, Zander M, Maier M. Tropheryma
whippelii endocarditis confirmed by polymerase chain reaction. Eur
Heart J 1995; 16:424-5.
70. Rickman LS, Freeman WR, Green WR, et al. Brief report: uveitis caused
by Tropheryma whippelii (Whipple's bacillus). N Engl J Med 1995;
332:363-6.
71. Wilson KH, Blitchington R, Frothingham R, Wilson JAP. Phylogeny of
the Whipple's disease-associated bacterium. Lancet 1991;338:474-
475.
72. Harmsen 0, Heesemann J, Brabletz T, Kirchner T, Muller-Hermelink
HK. Heterogeneity among Whipple's-disease-associated bacteria [let-
ter]. Lancet 1994;343:1288.
73. Sherlock SO. Diseases of the liver and biliary system. 9th ed. Oxford:
Blackwell, 1993:461.
157
74. Hopf U, Moller B, Stemerowicz R, et al. Relation between Escherichia
coli R (rough)-forms in gut, lipid A in liver, and primary biliary cirrho-
sis. Lancet 1989;2:1419-22.
75. Vilagut L, Vila J, Vinas 0, et al. Cross-reactivity of anti -Mycobacterium
gordonae antibodies with the major mitochondrial autoantigens in pri-
mary biliary cirrhosis. J HepatoI1994;21:673- 7.
76. James DG, ed. Sarcoidosis and other granulomatous disorders. New York:
Marcel-Dekker, 1994.
77. Mankiewicz E. Mycobacteriophages isolated from persons with tubercu-
lous and nontuberculous conditions. Nature, London, 1961: 191:1416
1417.
78. Mankiewicz E. The relationship of sarcoidosis to anonymous bacteria.
Acta Med Scand Suppl 1964;425:68- 73.
79. Chapman JS, Speight M. Further studies of mycobacterial antibodies in
the sera of sarcoidosis patients. Acta Med Scand Suppl 1964;425:
61-7.
80. Richter J, Bartak F, Halova R. Detection of mycobacteria by fluorescent
microscopy in sarcoidosis. In: Proceedings of the 5th International
Conference on Sarcoidosis. Prague: Universita Karlova, 1971:83-4.
81. Saboor SA, Johnson NM, McFadden J. Detection of mycobacterial DNA
in sarcoidosis and tuberculosis with polymerase chain reaction. Lancet
1992; 339: 1012-5.
82. Mitchell IC, Turk JL, Mitchell DN. Detection of mycobacterial rRNA in
sarcoidosis with liquid-phase hybridisation. Lancet 1992: 339:
1015-7.
83. Gerdes J, Richter E, Rusch-Gerdes S, et al. Mycobacterial nucleic acids
in sarcoid lesions [letter]. Lancet 1992; 339: 1536- 7.
84. Thakker B, Black M, Foulis AK. Mycobacterial nucleic acids in sarcoid
lesions [letter]. Lancet 1992;339:1537.
85. Bocart D, Lecossier 0, De Lassence A. Valeyre D, Battesti J-P, Hance
AJ. A search for mycobacterial DNA in granulomatous tissues from
patients with sarcoidosis using the polymerase chain reaction. Am Rev
Respir Dis 1992; 145:1142--8.
86. Scadding JG. Further observations on sarcoidosis associated with M.
tuberculosis infection. In: Proceedings of the 5th International Confer-
ence on Sarcoidosis. Prague: University Karlova, 1971:89-92.
87. Hanngren A, Biberfeldt G, Carlens E, et al. Is sarcoidosis due to an
infectious interaction between virus and mycobacterium? In: Proceed-
ings of the 6th International Conference on Sarcoidosis. Tokyo: Uni-
versity of Tokyo Press, 1974:8-11.
88. Nakata Y, Kataoka M, Kimura 1. Sarcoidosis and Proprionibacterium
acnes [in Japanese]. Nippon Rinsho 1994;52:1492-7.
89. Graham DY, Markesich DC, Kalter DC. Yoshimura HH. Isolation of
cell-wall defective acid-fast bacteria from skin lesions of patients with
sarcoidosis. In: Proceedings of the 11th World Congress on Sarcoidosis
and Other Granulomatous Disorders. Amsterdam: Excerpta Medica.
Elsevier, 1988: 1614.
90. Fidler HM, Rook GA, Johnson NM, McFadden J. Mvcobacterium tuber-
culosis DNA in tissue affected by sarcoidosis. BM.I 1993; 306:546- 9.
91. Reid .ID, Chiodini RJ. Serologic reactivity against Mvcobacterium para-
tuberculosis antigens in patients with sarcoidosis. Sarcoidosis 1993:
10:32-5.
92. Milman N, Andersen AB. Detection of antibodies in serum against M
tuberculosis using western blot technique. Comparison between sar-
coidosis patients and healthy subjects. Sarcoidosis 1993; 10:29 3 I.
93. Abe C, Iwai K, Mikami R. Hosoda Y. Frequent isolation of Propionibac-
terium acnes from sarcoidosis lymph nodes. Zentralbl Bakteriol Mik-
robiol Hyg (B) 1984;256:541-7.
94. Miyagawa Y, Asoh H, Nakanishi M, Shigematsu N. Epithelioid cell
granuloma formation in Lewis rats induced by streptococcal cell wall
(SCW). In: Abstracts of the 12th World Congress on Sarcoidosis and
Other Granulomatous Disorders. Kyoto, Japan: University of Tokyo
Press, 1989: I07.
95. Montemurro L, Rizzato G. Is sarcoidosis a borreliosis? Sarcoidosis 1991;
8:134-5.
158 Zumla and James
96. Bing H, Quing L, Wang F, Cheng A, Wei 1. Borrelia burgdorferi infec-
tion may be the cause of sarcoidosis. Chin Med J 1992;195:560-3.
97. HannukselaM, JannsonE. Isolationof a mycoplasmafrom sarcoidtissue.
In: Iwai K, HosodaY, eds. Proceedingsof the 6th InternationalConfer-
ence on Sarcoidosis. Tokyo: Tokyo University Press, 1974:4.
98. Uesake E, Izumi T, Tsuki S. Nocardia-like organisms isolated from le-
sions of sarcoidosis. In: Iwai K, Hosoda Y, eds. Proceedings of the
6th InternationalConferenceon Sarcoidosis. Tokyo: TokyoUniversity
Press, 1974:3.
99. RottoliP, BianchiBandinelliML, Rottoli L, Zazzi M, PanzardiG, Valen-
sin PE. Sarcoidosis and infections by human lymphotropic viruses.
Sarcoidosis 1990;7:31-3.
100. Hirshaut Y, Glade P, Vieira LO, Ainbender E, Dvorak B, Siltzbach LE.
Sarcoidosis, another disease associated with serologic evidence for
herpes-like virus infection. N Engl J Med 1970;283:502-6.
101. Ablashi D, SalahuddinZ, Imam P, et al. Serologicalassociationof a new
herpes virus (HBLV) with sarcoidosis. In: Proceedings of the II th
Congresson Sarcoidosisand Other Granulomatous Disorders.Amster-
dam: Excerpta Medica, 1988:68.
102. RottoliP, ValensinPE, Bianchi BandinelliM, BracialeP, FortelioniGM,
Vagliasindi M. Antibody responses to viral antigens in sarcoidosis
[letter]. Sarcoidosis 1993;10:73-4.
103. HillerdalG, Frisk G, Nettelbadt 0, DiderholmH. High frequencyofIgM
antibodies to coxsackie B virus in sarcoidosis patients and patients
with asbestos-related lesions. Sarcoidosis 1992;9:39-42.
104. Petterson HB, Iversen OJ, Eklung A. Retrovirus: a possible etiological
agent of sarcoidosis. In: Abstracts of the 12th World Congress on
Sarcoidosis and Other Granulomatous Disorders. Kyoto, Japan: Uni-
versity of Tokyo Press, 1989:120.
105. Shelley WB, Hurley HJ. The allergic origin of zirconium deodorant
granulomas. Br J Dermatoll958; 70:75.
106. CummingsMM,DunnerE, SchmidtBH, Barnwell JB. Conceptsof epide-
miology of sarcoidosis. Postgrad Med J 1956;19:437.
107. Cummings MM, Dunner E, Williams JH Jr. Epidemiologic and clinical
observations in sarcoidosis. Ann Intern Med 1959;50:879-90.
108. Comstock GW, Keltz H, Sencer DJ. Clay eating and sarcoidosis: a con-
trolled study in the state of Georgia. Am Rev Respir Dis 1961;84(5
part 2)(suppl):130-4.
109. Rubin E, Farber JL. The gastrointestinal tract. In: Rubin E, Farber JL,
eds. Essential pathology. Philadelphia: Lippincott, 1988.
110. WakefieldAJ, Sawyerr AM, Dhillon AP, et al. Pathogenesisof Crohn's
disease: multifocal gastrointestinal infarction. Lancet 1989;2:1057-
62.
Ill. Kayser K, Probst F, Gabius HJ, MOller KM. Are there characteristic
alterations of lung tissue associated with Crohn's disease? Pathol Res
Pract 1990;186:485-90.
112. Wakefield AJ, Pittilo RM, Sim R, Cosby SL, Stephenson JR, Dhillon
AP, Pounder RE. Evidence of persistent measles virus infection in
Crohn's disease. J Med ViroI1993;39:345-53.
113. SandersonJD. Mycobacteria in Crohn's disease [letter]. BMJ 1993;306:
1131.
ern 1996;23 (July)
114. Ciclitira PI. Does Crohn's diseasehave a mycobacterialbasis? [editorial].
BMJ 1993;306:733-4.
115. Smith MSH, WakefieldAJ. Viral association with Crohn's disease. Ann
Med 1993;25:557-61.
116. EkbomA, WakefieldAJ, Zack M, Adami HO. Perinatalmeaslesinfection
and subsequent Crohn's disease. Lancet 1994;344:508-10.
117. Lewin J, Dhillon AP, Sim R, Mazure G, Pounder RE, Wakefield AI.
Persistent measles virus infection of the intestine: confirmation by
immunogold electron microscopy. Gut 1995;36:564-9.
118. WakefieldAJ, Ekbom A, Dhillon AP, Pittilo RM, Pounder RE. Crohn's
disease: pathogenesis and persistent measles virus infection. Gaster-
oenterology 1995; 108:911-6.
119. James DG. Definition and classification. In: James DG, ed. Sarcoidosis
and other granulomatous disorders. New York: Marcel-Dekker, 1994:
33-5.
120. Hance AJ, CandranelJ, Soler P, Basset F. Pulmonaryand extrapulmonary
Langerhans' cell granulomatosis (Histiocytosis X). Semin Respir Med
1988;9:349-68.
121. Smith RM, Curnutte JT. Molecular basis of chronic granulomatousdis-
ease. Blood 1991;77:673-86.
122. Curnutte JT. Chronic granulomatous disease: the solving of a clinical
riddle at the molecular level. Clin Immunol Immunopathol 1993;
67(3)(suppl):S2-15.
123. Thrasher AJ, Keep NH, Wientjes F, Segal AW. Chronic granulomatous
disease. Biochim Biophys Acta 1994;1227:1-24.
124. SchapiroBL, NewburgerPE, KlempnerMS, DinauerMe. Chronicgranu-
lomatous disease presenting in a 69-year-old man. New Engl J Med
1991;325:1786-90.
125. Melkersson E. Case of recurrent facial paralysis with angioneurotic
edema. Hygiea 1928;90:737-41.
126. RosenthalC. Facialislahmungund lingua plicata.Z Neurol Psycholl990;
131:475-501.
127. Larsson E, Westermark P. Chronic hypertrophic vulvitis-a condition
with similarities to cheilitis granulomatosa (Melkersson-Rosenthal
syndrome). Acta Dermatovenereol (Stockholm) 1978;58:92-3.
128. Klaus SN, Brunsting LA. Melkersson's syndrome(persistentswelling of
the face, recurrent facial paralysis and lingua plicata): report of case.
Proc Mayo Clin 1959;34:365- 70.
129. WiesenfeldD, FergusonMM, MitchellDN, et al. Oro-facialgranulomato-
sis: a clinical and pathological analysis. Q J Med 1985;54:101-113.
130. Kikuchi M. Lymphadenitis showing focal reticulumcell hyperplasiawith
nuclear debris and phagocytosis. Nipon Ketsueki Gakkai Zassi 1972;
35:379-80.
131. Kuo T-t. Kikuchi's disease (histiocyticnecrotizinglymphadenitis): a clin-
icopathologic study of79 cases with an analysis of histologicsubtypes,
immunohistology, and DNA ploidy. Am J Surg Patho11995;19:798-
809.
132. Gourley I, Bell AL, Biggart D. Kikuchi's disease as a presenting feature
of mixed connective tissue disease. Clin Rheumatoll995; 14:104-7.
133. Fred HL. Kikuchi-Fujimoto disease. Houston Med 1991;7:115-7.