ChapterPDF Available

Legionella and legionella-like organisms



Medicine is an ever-changing science. As new research and clinical experience broaden our knowledge, changes in treatment and drug therapy are required. The author and the publisher of this work have made very effort to ensure that the treatment and drug dosage schedules herein are accurate and in accord with the standards accepted at the time of publication. Readers are advised, however , to check the product information sheet in­ cluded in the package of each drug they plan to administer to be certain that changes have not been made in the recommended dose or in the indications and contraindications for administration. This recommendation is of particular importance in regard to new or infre­ quently used drugs.
Clinical Approaches to Infectious
Diseases of the Lower Respiratory Tract
Edited by
Matthew E. Levison
Iohn Wright . PSG Inc
Boston Bristol London
Library of Congress in Publication Data
Main entry under title:
The Pneumonias : clinical approaches to infectious
diseases of the lower respiratory tract.
Bibliography: p.
Includes index.
l. Pneumonia. I. Levison, Matthew E., 1938-
IDNLM: l. Pneumonia. WC 202 P7311
RC771.P685 1984
ISBN 0-7236-7020-X 616.2'41 83-10473
Published simultaneously by:
John Wright. PSG Inc, 545 Great
Massachusetts 01460, U.S.A.
John Wright & Sons Ltd,823-825
Bristol BS4 sNU, England
Road, Littleton,
Bath Road,
Medicine is an ever-changing science. As new research and clinical experience broaden our
knowledge, changes in treatment and drug therapy are required. The author and the
publisher of this work have made very effort to ensure that the treatment and drug dosage
schedules herein are accurate and in accord with the standards accepted at the time of
publication. Readers are advised, however, to check the product information sheet in-
cluded in the package of each drug they plan to administer to be certain that changes have
not been made in the recommended dose or in the indications and contraindications for
administration. This recommendation is of particular importance in regard to new or infre-
quently used drugs.
Copyright O 1984 by John Wright . PSG Inc
All rights reserved. No part of this publication may be reproduced or transmitted
in any form or by any means, electronic or mechanical, including photocopy'
recording, or any information storage or retrieval system, without permission in
writing from the publisher.
Printed in the United States of America.
International Standard Book Numb er : 0-7 236-7020-X
Library of Congress Catalog Card Number: 83-10473
Elias Abrutyn, MD
Professor and Assistant Chairman
Department of Medicine
Division of Infectious Diseases
The Medical College of Pennsylvania
Assistant Chief of Medicine and
Chief, Infectious Diseases
V.A. Hospital of Philadelphia
Philadelphia, Pennsylvania
Stephen Baumgart, MD
Assistant Professor of Pediatrics
The University of Pennsylvania
School of Medicine
Philadelphia, Pennsylvania
Jerome Boscia, MD
Assistant Professor of Medicine
Division of Infectious Diseases
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Jaime Carrizosa, MD,
Altamonte Springs, Florida
(Formerly with the Division of
Infectious Diseases,
The Medical College of Pennsylvania)
Paul E. Epstein, MD
Associate Professor of Medicine
The University of Pennsylvania
Chief, Pulmonary Medicine
Graduate Hospital
Philadelphia, Pennsylvania
Harvey M. Friedman, MD
Associate Professor of Medicine
The University of Pennsylvania
Director, Diagnostic Virology
Children's Hospital
Philadelphia, Pennsylvania
Richard Gilpin, PhD
Associate Professor of Microbiology
and Immunology
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Linda Griska, MD
Associate Professor
Department of Radiology
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Donald Kaye, MD
Professor and Chairman
Department of Medicine
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Oksana M. Korzeniowski, MD
Assistant Professor of Medicine
Division of Infectious Diseases
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Jack L. LeFrock, MD
Professor of Medicine and Chief
Division of Infectious Diseases and
Clinical Microbiology
Hahnemann University School of
Philadelphia, Pennsylvania
Arnold Lentnek, MD
Group Director, Anti-Infective
Medical Affairs Department
Smith Kline & French Laboratories
Matthew E. Levison, MD
Professor of Medicine and Chief
Division of Infectious Diseases
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Bennett Lorber, MD
Professor of Medicine
Chief, Section of Infectious Diseases
Temple University Health Sciences
Philadelphia, Pennsylvania
Rob Roy MacGregor, MD
Associate Professor of Medicine
Chief of the Infectious Diseases
The University of Pennsylvania
School of Medicine
Philadelphia, Pennsylvania
Paul L. Marino, MD
Assistant Professor of Medicine
Graduate Hospital and Hospital of
the University of Pennsylvania
Philadelphia, Pennsylvania
Richard V. McCloskey, MD
Associate Director
Department of Medical Research
Hoffman-LaRoche, Inc.
Nutley, New Jersey
William B. McNamee, Jr., MD
Director of Infectious Diseases
Mercy Catholic Medical Center
Philadelphia, Pennsylvania
(Formerly with the Division of
Infectious Diseases
The Medical College of Pennsylvania)
Abdolghader Molavi, MD
Associate Professor of Medicine
Division of Infectious Diseases and
Clinical Microbiology
Hahnemann University School of
Philadelphia, Pennsylvania
Sheila Ann Murphey, MD
Clinical Associate Professor of
Director, Division of Infectious
Jefferson Medical College of
Thomas Jefferson University
Philadelphia, Pennsylvania
Richard A. Polin, MD
Assistant Professor of Pediatrics
University of Pennsylvania
School of Medicine
The Children's Hospital of Philadelphia
Philadelphia, Pennsylvania
George A. Poporad, MD
Clinical Assistant Professor of
The Medical College of Pennsylvania
Co-Head of the Section of Infectious
Frankford Hospital
Philadelphia, Pennsylvania
(Formerly with the Division of
Infectious Diseases
The Medical College of Pennsylvania)
Jerome Santoro, MD
Clinical Associate Professor of
Jefferson Medical College of Thomas
Jefferson University
Philadelphia, Pennsylvania
(Formerly with the Division of
Infectious Diseases
The Medical College of Pennsylvania)
Richard Snepar, MD
Clinical Instructor of Medicine
Rutgers Medical School-U.M.D.N. J.
Piscataway, New Jersey
(Formerly with the Division of
Infectious Diseases
The Medical College of Pennsylvania)
Jack D. Sobel, MD
Associate Professor of Medicine
Division of Infectious Diseases
The Medical College of Pennsylvania
Philadelphia, Pennsylvania
Alan R. Spitzer, MD
Assistant Professor of Pediatrics
and Obstetrics and GynecologY
University of Pennsylvania
School of Medicine
Philadelphia, Pennsylvania
George H. Talbot, MD
Assistant Professor of Medicine
Hospital Epidemiologist
The Hospital of the University of
Philadelphia, Pennsylvania
Clifford G. Wodaver, MD
Clinical Instructor in Medicine
at the University of Oklahoma
Health Science Center
College of Medicine
Oklahoma City, Oklahoma
Preface xi
I eathogenesis of Pneumonia I
Matthew E. Levison, Jerome Boscia
2 Host Defenses in the Respiratory Tract E
Sheila Ann Murphey
3 Radtographic Manifestations of Pneumonia: An Approach to
Dffirential Diagnosis 23
Linda Griska
4 Complications of Pneumonia 39
Paul L. Marino
5 Use of the Clinical Microbiology Laboratory 47
Matthew E. Levison
6 (Jse of Antimiuobiol Agents 66
Matthew E. Levison
7 Pulmonary Function Defects in Lower Respiratory Troct
Infection 9l
-. Paul E. Epstein
'8 Respiratoiy Care in Pneumonia 100
Paul E. Epstein
, 9 Strategies in the Approach to Treatment of pneumonia 110
Elias Abrutvn
l0 Typical Pneumonia Syndrome lZl
Oksana M. Korzeniowski
I I Atypicat Pneumonia Syndrome 130
Jerome Boscia, Oksana M. Korzeniowski
12 Chronic Pneumonias 137
Jaime Carrizosa, William B. McNamee, Jr.
13 Recurrent Pneumonio 153
Elias Abrutvn
14 Pneumoniai of the Newborn Period and Infancy l6j
< F Stephen Baumgart, Alan R. Spitzer, Richard A. polin
I ) Nosocomial Respiratory Tract Infections 182
Jerome Santoro vll
Pneumonia in the Aspiration-Prone Pqtient 197
Bennett Lorber
Pulmonary Infections in the Immunocompromised Host 206
Jack D. Sobel
Pneumonia in the Tronsplant Patient 242
George H. Talbot
19 Streptococcus pneumoniae 261
_ - 1 Arnold Lentnek, Jack L. LeFrock, Abdolghader Molavi
?g' Pneumococcal Pneumonia Prevention 272
- Richard V. McCloskev
2l Streptococcw pyogrrL, (Group A Streptococci) 278
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
22 Group B Streptococci 2E2
A^ Alan R. Spitzer, Stephen Baumgart, Richard A. Polin
Z5 Staphylococcus eureus 290
Arnold Lentnek, Jack L. LeFrock, Abdolghader Molavi
24 Bocillus anthracis and Other Bacillus Species 301
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
25 Nocardia Species 305
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
26 Enterobocteriaceae, Pseudomonas aeruginosa ond
Acinetobacter 309
Abdolghader Molavi, Jack L. LeFrock
Pseudomonas pseudomollei 335
Jack L. LeFrock, Arnold Lentnek, Abdolghader Molavi
Pseudomonas mallei 338
Matthew E. Levison
Yersinia pestis 340
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
Froncisella tularensis 343
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
Hemophilus influenzae 347
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
Bordetellapertussis 352
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
Neisseria meningitidis 355
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
Legionella and Legionella-like Organisms 358
Richard Gilpin
Other Obligate Anaerobes Including Actinomyces 372
Jack L. LeFrock, Abdolghader Molavi, Arnold Lentnek
Mycobacteria 390
Clifford G. Wlodaver, Rob Roy MacGregor
Mycoplasma pneumoniae 418
Richard Snepar
Branhamella catarrhalis 422
Matthew E. Levison
Chlomydiae 424
George A. Poporad
Coxiella burnetii (Q-Fever) 431
George A. Poporad
4l Respiratory Syncytial Virus 435
Richard Snepar
42 Rhinovirus 43E
Richard Snepar
43 Measles 441
Richard Snepar
44 Coxsockievirus 445
George A. Poporad
45 Varicella-Zoster 447
George A. Poporad
46 Herpes Simplex 453
George A. Poporad
47 Epstein-Barr Virus 461
George A. Poporad
48 Adenovirus 467
Richard Snepar
49 Cytumegolovirus 471
Clifford G. Wlodaver, Harvey M. Friedman
50 Influenza 475
George A. Poporad
51 Aspergillus 483
George H. Talbot
52 Mucorales 498
George H. Talbot
53 Candida 502
Ceorge H. Talbot
54 Histoplasma capsulatum 506
Jaime Carrizosa
55 Blastomyces dermatitidis 510
Jaime Carrizosa
56 Cryptococcus neoformans 513
Jaime Carrizosa
57 Coccidioides immitis 516
Jaime Carrizosa
58 Paracoccidioidesbrasiliensis 520
Matthew E. Levison
Protozoa and Helminths
59 Protozoo and Helminths SZI
Donald Kaye
Index 556
. A Legionella snd
J+ I-egionellq-like Organisms
Richard W. Gilpin
/Legionetla and, Legionella-like organisms (LLO) are gram-negative
bacilli which have a fairly uniform width (0.3 to 0.9 pm) and length (2'5
pm). A few bacteria in some isolates are much longer. The bacteria usually
are found as individual cells, but may form chains or long filaments'
Legionella and LLO have an outer and inner membrane structure typical
of gram-negative bacteria (Figure 34-l)/ Electron lucent cytoplasmic
vacuoles, which probably contain poly-0-hydroxybutyrate' are often
observed (Figure 34-l). These vacuoles appe,ar as large bumps within
Legionella in scanning electron micrographs.4hese bacteria do not form
.ndorpor.s and are not encapsulated. Legionellohave a branched-chain
fatty acid composition that is unusual for gram-negative bacteria and
affords identification on the .basis of analysis of cellular fatty acids by
gas-liquid chromatog raPhY /
- A single polar flagellum has been found on isolates that have been
grown on media that permits ra-pid growth. Flagellated Legionella have
also been observed in human lung tissue in areas that do not have a dense
population of Legionella or do not have a heavy infiltrate of inflam-
*utoty cell debris. Pili (fimbrae), one half the diameter of the flagellum,
have also been observed by electron microscopic studies of Legionella
grown in vitro.
- ,/There are no unique morphological characteristics that can easily
distinguish Legionello from other gram-negative rcds/The cytoplasmic
uu.rrol.r, flagellum, and pili are not always observable and may be miss-
ing in some species.
Staining Characteristics
/Leptonetla and LLO from in vitro cultures all stain gram negative'
The piik color may be faint, however, unless the safranin or carbol
fuchsin (preferred) counterstain is left on smears for at least one minute'
fegioneiia in tissue or other clinical specimens will usually not take the
g.u- stain. Therefore, the gram stain procedure has limited usefulness
ior identification of these bacteria in clinical material. Legionella and
Figure 34-L Legionella pneumophila structure in thin section. The Legionella
were fixed with glutaraldehyde, followed by osmium tetroxide, dehydrated in
ethanol, embedded in Epon and sectioned. Sections were stained with uranyl
acetate and lead citrate. The wrinkled cell wall outer membrane and the inner
cytoplasmic membrane are quite apparent. Two electron-lucent cytoplasmic in-
clusions can be seen. Magniflcation: 140,000 X.
LLO are not considered to be acid-fast/
There are two staining procedures that may identify bacteria such as
Legionella in lung or other tissues. Various modifications of the Giemsa
stain have produced variable results because of differentiating Legionella
from a background of inflammatory cell debris. The Dieterle silver im-
pregnation stain originally developed for staining spirochetes in tissue is
useful, because the bacteria appear as dark red rods against a yellow to
pink background of cellular debris. The Warthin-Starry silver impregna-
tion method also produces satisfactory results. Other standard histo-
logical stains, including hematoxylin and eosin, are rarely useful.
Flagella may be stained by the Leifson flagellar stain but this has
limited application for clinical diagnosis. B-hydroxybutyrate inclusions
of Legionella may stain blue-black with the Sudan Black B fat stain, but
this also has limited diagnostic usefulness.
Growth Characteristics
/Legionella and LLO are unusual because they do not grow on the
routine bacteriological media available in clinical microbiology labora-
tories. This includes sheep blood agar, chocolate agar, nutrient agar,
Mueller-Hinton agar, trypticase soy agar, and all of the selective/differential
agar or broth media used for gram-negative enteric bacteria/There is
some evidence to suggest that peroxides in the above culture media may
inhibit the growth of Legionella. For this reason, the types of media for
Legionella culture will be reviewed. The first in vitro culture medium to
be developed utilized a Mueller-Hinton agar base, modified by the addi-
tion of 190 Isovitalex and I go hemoglobin. However, Legionella growth
on this medium is slow. It was soon discovered that these bacteria grow
better on a medium containing casein hydrolysate plus beef extractives or
yeast extract when two filter-sterilized supplements, L-cysteine and ferric
pyrophosphate-soluble were added./these media are known as Feeley-
Gorman (FG) agar and charcoal (0.2v0 activated charcoal) yeast extract
(CYE) agar, respectivelV.gSome Legionella do not grow well on FG agar.
Therefore, the current medium of choice is cYE agar modified by the in-
clusion of a buffer. The most commonly used filter sterilized broth
medium is the same as CYE agar with the omission of activated charcoal
and agar. Autoclaved broth medium may not support growth.
Further modifications of CYE agar have been made in order to
differentiate between the various species of Legionella andinhibit growth
of other bacteria. Buffered CYE agar medium containing glycine,
a-ketoglutarate, cefamandole, polymyxin B, anisomycin, vancomycin,
bromthymol blue, and bromcresol purple may permit isolation of
Legionella from samples that contain high numbers of other bacteria.
However, buffered cYE agar remains the first choice for primary isola-
tion of Legionella and LLo. Stock cultures can be stored in the dark at
room temperature on CYE agar slants for at least two months.
Legionella grow best on CYE medium in humid air supplemented
with2.5Vo COz at 35'C. Primary cultures from clinical sources may re-
quire five to seven days before colonies can be seen. Unfortunately, other
bacteria present in a specimen may overgrow and/or inhibit the growth
of Legionella during this prolonged incubation period. Primary clinical
or environmental samples may need to be passed through guinea pigs
and/ar embryonated.hen eggs before Legionella or LLO can be isolated
on CYE agar. Legionella and LLO are aerobic bacteria that utilize amino
acids as a source of carbon and energy. However, Legionella may use
some carbohydrates as a carbon source.
Speciation of Legionella and LLO has moved rapidly during the last
few years. There are currently nine proposed species; L. pneumophilo, L.
micdadei, L. bozemanii, L. dumoffii, L. gormanii, L. longbeachae, L.
wadsworthii, L. jordanis, andL. oakridgensls. There are additional LLO
isolates that have not been classified at this time. There seems to be a
precedent for retaining the genus name, Legionella, for the isolates that
have been characterized. The current literature also uses other taxonomic
names, original isolate numbers or letter designations, and even different
genus names. Therefore, the following tables have been compiled to
show the taxonomic (Table 34-l) and physiological (Table 34-2)
differences among the Legionella that have been investigated.
There are currently eight serogroups of L. pneumophila (Table
34-3). Most of the sporadic clinical cases of Legionnaires' disease studied
through 1979 were caused by L. pneumophila, serogroup 1, followed by
serogroups 2, 3 and 4. Data is incomplete on the frequency of serogroups
5 and 6 among clinical cases. There appears to be a Legionella common
antigen, termed F-1, which is located on the surface of the bacterium.
The genome size of Legionella is in the range of 2.5 x lOe daltons
and the guanine plus cytosine (GC) content is between 3590 and 43V0,
with an average content of 39V0. No genetic relationship to previously
known bacteria has been established by DNA hybridization studies.
Plasmids have been detected in some Legionello isolates, but what they
code for is not known.
Table 34-1
Taxonomic Classification of Legionella and LLO
Proposed Name Original Isolate
Designation* Other Proposed
Legione I la pneumop hi la
L. micdadei
L. bozemanii
L. dumoffii
L. gormanii
L. longbeachae
L. jordanis
L. oakridgensis
L. wadsworthii
Legionnaires' disease
bacillus, Philadelphia,
Pittsburgh pneumonia
agent (PPA)
NY-23, Tex-Kl
Long Beach 4
Oak Ridge I0
Wqdsworth 8l-7164
Tatlockia micdadei
L. pittsburghensis
Fluoribacter bozemanae
Fluoribacter dumffii
Fluoribacter gormanii
*Italicized original isolate designation is the type strain for that species.
Table 34-3
Legionella pneumophila Serogroup Strains
Bacterial Isolate Serogroup
Knoxville I
Togus I
Bloomington 2
Los Angeles I
Dallas lE
Chicago 2
Chicago 8
Concord 3
Distribution in Nature
iLegionello and LLO are distributed ubiquitously in natural aquatic
environments. Legionella are also found in man-made aquatic en-
vironments such as shower heads, faucets, hot water tanks, heat rejec-
tion devices like cooling towers and evaporative condensers used for air
conditioning, and, most recently, recreational whirlpools. There does
not seem to be a human or animal reservoir. Cases of legionellosis have
been found in most areas of the United States and in many countries
around the world;
Epidemiologic investigations of legionellosis have been based on
retrospective evaluations of patient sera and postmortem tissue. In
studies of sera from adult populations, approximately 4vo to 590 of the
people have antibody titers to serogroup I I. pneumophila which sug-
gests possible past infection but not necessarily disease.
A retrospective study by England et al (1981) of the flrst 1005 cases
of documented sporadic legionellosis caused by L. pneumophila
serogroups l-4 produced many interesting results. Cases were reported
in all states, except Alaska and South Dakota. The attack rate in 1978
was 2.4 cases per million population. Ages ranged from 16 months to 89
years with a median age of 55 years for males and 56 years for females.
,flh, relative risk for males was 2.6-fold higher than for females. Currently,
250,000 cases of legionellosis are estimated to occur each year in the
United States. Sporadic cases seem to be more common than outbreaks.
The epidemiological relationships to age, underlying disease, immuno-
suppression, smoking, travel, and exposure to. recent excavations are
similar for both sporadic cases and outbreaks./
Sporadic cases and outbreaks of legionellosis in the United States
and in other countries occur more frequently during the summer months.,4
I rYl
I lJr
t€ -.i
lc'\ Y
l'ri .O
| -@
la Y
tE 5
l* q
; 9, E E,aE
s Ee- .sq6e E
Frp'45-J eHiE!,eHEE
i.YE 5EE+€g€3 g E E EE; g ;EE
2Y- - Fp eSEEE fJE H Rx $ I ac d
E* 6€# -S5558fl'frz#E -t'
IJ() -
l-1 c)
7\ 6)
ca -=
-b0 6
Eq.. o
. vA qrrl
F"hu x>
\/ lJr
I l$l+ * r* I++ | tZZ
I r+ r++ | t22
++ |
++ |
++ |
al n
' qF
Z-'f | -f tl I I ZZ
TZ S r r
-\ 'E
T'z lva I T
rH= |
-z bi' -
l+cd l+
r r HH
+ t+ t++t I t+ |
l+sr+ 2++ t+t+tt l+t+
t-L I
L/ E{ .O
+r+Rr+ ++l+t+++r+t I
Isolation of L. pneumophila from natural lakes or ponds in the United
States is highest during the months of May, June and July. Of 1005
sporadic cases during the 1977-1979 period, 7390 occurred between the
months of June and October, with most cases (43V0) during the months
of August and September.
Outbreaks of legionellosis have often been related to airborne
transmission. The current evidence implicates the moist air exhaust of
cooling towers, evaporative condensers, and other air-handling equip-
ment as possible sources. Although L. pneumophila are often identified
in concentrated water samples from cooling towers and evaporative con-
densers, the bacterium has not been isolated directly from the aerosol ex-
haust of air-conditioning equipment. Laboratory tests of the viability of
L. pneumophila in aerosols indicate that the relatively short half-life of
its viability is greater in humid air.
If cooling towers are a source of infective Legionella, there should
be an increased risk for people who service or maintain the equipment.
However, a serologic survey of mechanics, mechanic's helpers, and water
treatment servicemen who maintain 3000 cooling towers in the New York
metropolitan area failed to confirm this.
Legionella have been isolated from shower heads in hospitals and
hotels. However, acquisition of legionellosis from these sources has not
been proven conclusively.
Yirulence Factors
,4fn, factor or factors responsible for the ability of Legionella to
establish the disease state have yet to be identified...An early hypothesis
involved endotoxin activity beca.use most gram-negative rods have the
lipid A-associated endotoxin in the lipopolysaccharide of their cell
wall. Legionella do produce a strongly positive Limulus lysate test, a
nonspecific in vitro test for endotoxin. However, Legionella do not pro-
duce significantly positive results in the more specific in vivo rabbit
pyrogenicity and mouse lethality tests.
:-"Exotoxins and,/or exoenzymes have been suggested as possible
virulence factors and include proteolytic, hemolytic and cytotoxic ac-
tivities. Numerous other enzymes, including an enzyme that inhibits
oxygen-dependent killing by phagocytic cells have been described in the
Legionella do not seem to be highly infective once they have been
passed several times in vitro. Passage of in vitro grown Legionella
through hen embryonated yolk sac or guinea pigs restores a moderate
level of virulence.
Investigation of the infectivity of Legionella aerosols for guinea pigs
found that the geometric mean LDso dose was approximately 3 x lff
bacteria per lung (Davis et al, 1982). The lethal dose by the intranasal or
intraperitoneal routes is reportedly much greater. It is apparent from
other reports that guinea pigs may be infected by a relatively low aerosol
dose (100 to 200 bacteria) without developing lethal disease. Also, no
cross-infection to a sentinal guinea pig placed in a cage with an infected
guinea pig has been found. The infective dose for the healthy human host
has not been determined. However, patients who have a compromised
immune response are possibly at higher risk. Only one case of possible
patient-to-patient transmission has been reported. However, the CDC
recommends that known cases be placed in respiratory isolation.
Guinea pig peritoneal macrophages isolated from normal guinea
pigs phago cytize yolk sac-grown Philadelphia 1 strain of L. pneumophila
at a lower rate than do macrophages isolated from previously infected,
convalescing guinea pigs. Adding immune sera increases the rate of
phagocytosis by macrophages from both normal and immune guinea
pigs. However, once Legionella are phagocytized, they begin to multiply
in vacuoles and eventually kill the macrophages from both normal and
convalescing guinea pigs. This is not influenced by the presence of im-
mune Sera. Legionella grown in vitro are also phagocytized, but they are
subsequently killed by the macrophages. Legionella within histiocytes or
macrophages in tissue from both humans and guinea pigs are often
observed to be intact with no evidence of bacterial lysis. ,There are un-
doubtedly some virulence-associated factors involved in the survival and
multiplication of Legionella within macrophages, but they remain to be
/ Legionellosis or Legionnaires' disease, whether occurring sporadically
or in an outbreak, is basically an acute bacterial bronchopneumonia. A
mild, nonpneumonic disease produced by Le7ionello has also been
described (Pontiac fever),'
The clinic4l aspects of severe legionellosis have been described by
many authors/The incubation period ranges from two to 10 days. Pneu-
monia is accompanied by a high fever (39o C-41' C); nonproductive
cough which may later become productive and possibly contain blood;
headache; a change in mental status; nausea; diarrhea; and vomiting/
Chest roentgenograms may initially show an infiltrate, usually confined
to a lower lobe, which may progress to bilateral involvement. Pulmonary
cavitation has also been reported. Renal and hepatic involvement have
been frequently described. Legionella antigen has also been found in
urine. Elevated serum lactic acid dehydrogenase, scor, and alkaline
phosphatase are frequently reported. Initial leukocyte counts among
sporadic cases vary from 4700 to 28,000 with an average of 9900/mm3.
The red cell sedimentation rate may become elevated (> 55 mm/hr) in
some cases.
Pathologic findings from fatal cases show a high frequency of
bilateral pulmonary consolidation. There is often a dense fibropurulent
pneumonia with bacteria at the edge of consolidated areas. Focal areas
of hemorrhage are common. There is an acute inflammatory reaction,
diffuse alveolar damage with intact Legionella present within alveolar
macrophages. Immunocompromised patients may have a fibroserous
pneumonia without much inflammation.
It is apparent that Legionella may spread by endobronchial, intersti-
tial, endolymphatic, and hematogenous routes. Bacteria may spread with
migrating phagocytic cells. Legionella have been found in hilar lymph
nodes, spleen, bone marrow, kidneys, and pulmonary venules.
Direct Fluorescent Antibody (DFA) Test
'/rnis test is used for the early diagnosis of legionellosis/specific
antisera are produced by immunizing rabbits with a formalinized or heat-
killed vaccine without Freunds' adjuvant. The gammaglobulin fraction
of the serum is conjugated with fluorescein isothiocyanate and used to
idenitfy Legionella in specimens by means of fluorescent microscopy.
Rhodamine-conjugated normal rabbit serum is used as a diluent in order
to counterstain autofluorescent debris. DFA reagents for L. pneumophilo
serogroups 1-6, L. bozemanii, L. dumofi, L. gormonii, L. micdadei,
and L. longbeachioe are now available from commercial sources-/ihe
DFA test is more specific and sensitive than the usual histologic staining
procedures. It can be done on sputum, if available, or on other respira-
tory secretions, such as tracheal aspirates, or pleural fluid,rL,iquifaction
and concentration of sputum may be helpful. This test can also be per-
formed on microscope slide imprints of fresh or formalinized lung tissue
or deparaffinized lung tissue sections. False-negative DFA tests may
occur as a result of inadequate numbers of Legionella in the specimens.
The omission of Freunds' complete adjuvant from the vaccines and
the inclusion of the rhodamine counterstain have eliminated much of the
nonspecific staining associated with some of the early DFA tests. A re-
cent report suggested, however, that patients with Bacteroides fragilis in-
fection may produce a positive, cross-reactive DFA test to serogroup I z.
pneumophilabut not to serogroups 2-4. Also, Pseudomonasfluorescens,
P. alcaligenes, and the Flavobacterium-Xanthomonas group were
reported to stain positively with serogroup I antisera.
Preparation of specimens, examination of slides, and interpretation
of DFA results are presented in the CDC Manual. Since there are very
few organisms in lung exudates, observation of > 5 fluorescing bacteria
is considered positive. For clinical specimens, except lung exudates, > 25
strongly fluorescing bacteria per smear is considered positive. If there are
< 25 strongly fluorescing bacteria per smear, the actual numbers are
reported. No strongly fluorescing bacteria per smear is reported as
negative. Lung exudates are difficult to evaluate because of the presence
of autofluorescing debris in the specimens. Therefore, familiarity with
the morphology of Legionella and experience with fluorescent micros-
copy are needed for proper interpretation. Appropriate controls must be
run in parallel with test samples.
,/ lt is recommended that nonformalinized specimens examined by the
DFA test also be cultured on buffered CYE agar with and without addi-
tional supplements. The combination of culture plus DFA test produces
results that are more diagnostically useful than when each is used alone.
A survey of sporadic cases found that nearly half of DFA-positive speci-
mens were culture positiver/ Blood should also be cultured since
Legionella have been isolated from the blood of a few patients with
legionellosis. A diphasic (broth/agar) CYE blood culture medium is
recommended but standard aerobic blood culture medium also has been
used successfully.
There are two reasons that may explain false-negative cultures in pa-
tients with legionellosis. First, viable Legionella may not be present in
respiratory secretions unless retrograde spread of the organism has oc-
curred in the airway. Second, human anterior pharyngeal flora contam-
inating specimens of respiratory secretions may inhibit the growth of
Legionella. For example, we have found that some isolates of
Staphylococcus aureus, Streptococcus pneumoniae, Pseudomonas
aeruginosa, and some viridans streptococci inhibit the growth of I.
pneumophila Philadelphia I on buffered CYE agar. Culture of clinical
specimens is recommended, however, because there have been reports of
occasional positive cultures from specimens that were DFA-negative.
Various modifications of the culture medium are being evaluated at
this time. Some of these modifications are designed to permit differentia-
tion between some species of Legionella, while others contain antibiotics
and other supplements to inhibit the growth of other bacteria and fungi
while permiting the growth of Legionella. At this time, buffered CYE
agar medium is recommended for primary isolation and is commercially
Laboratories with suitable biological containment facilities have
successfully recovered Legionella by intraperitoneal injection of clinical
specimens into guinea pigs. Culture of guinea pig spleen homogenates or
peritoneal washings after three to five days directly on CYE agar or on
CYE agar after passage through the yolk sac of embryonated hen eggs
has often resulted in isolation of Legionella when direct plating results
were negative. The procedures for in vivo passage are presented in the
CDC Manual.
Indirect Immunofluorescence (IFA) Test
.'This is a serological test for antibody to Legionella in patient's
serum. The IFA test can be used to confirm legionellosis when paired
acute (within seven days of onset) and convalescent (20 to 40 days after
onset) patient sera show a fourfold or greater rise in titer to > 128. When
only one serum sample is available, a single or standing titer of > 256 is
presumptive evidence of previous or current infection'7'
lfatients with legionellosis seroconvert to a diaghostically positive
IFA titer, unless they succumb to the disease before an antibody response
is elicited or they have a defective immune response ro the organism.
However, seroconversion may take up to three weeks. Therefore, the
IFA test has limited usefulness in the early diagnosis of legionellosis. The
DFA test and culture produce positive diagnosic results earlier, but they
may often be negative. In this instance, early diagnosis must be based on
clinical signs and symptoms and later confirmed by the IFA test"ss
The IFA test procedure and interpretation of results is presented in
the CDC Manual. State public health laboratories are equipped to per-
form this test. The antigens for the IFA test are Legionella that are
grown on CYE agar and heat-killed. The bacteria are suspended in nor-
mal chicken yolk sac fluid and fixed to a microscope slide. various dilu-
tions of patient's serum are incubated with the antigen. The slide is then
washed and incubated with fluorescein-conjugated antihuman serum to
determine whether antibodies in the patient's serum attached to the
known antigen. Fluorescence intensities are graded from 0 to 4+. The
highest dilution of the patient's serum that produces at least 1+ , barely
visible staining, is considered the end point. It is very important to
include appropriate controls with all tests. As with the DFA test, some
experience is required to evaluate staining intensities and background
staining of yolk sac material, filamentous bacteria and other artifacts.
The current commercially available IFA test kits using individual
serogroup bacterial preparations seem to have a sensitivity of 8590 and a
specificity of 92V0.
There are a variety of cross-reactions that may occur, however. Pa-
tient sera with antibody to other gram-negative bacteria may cross-react
with Legionella. This cross-reactivity may be effectively blocked with im-
munoabsorbant from Escerichia coli strain 013:K92:H4. However. this
procedure is not routinely used. No cross-reactions against L. pneumo-
phila antigens have been found in sera from cases of Q fever, tularemia,
or psittacosis, or in sera of patients with rheumatoid factor.
Other Diagnostic Tests
Several other tests for Legionella antigen and antibody have been
developed and compared to the more established DFA and IFA tests. A
brief description of these tests will suffice because most are not widely
used. A microagglutination test has been developed for detection of IgM
class antibody. A microenzyme-linked immunosorbent assay (ELISA)
has been developed to detect antlbodies to serogroup 1 l. pneumophilia
and Legionello antigen in urin#Various agglutination tests, indirect he-
magglutination, ELISA, radioimmunoassays (RIA), and counter-
immunoelectrophoresis tests to detect Legionella antigen have been
described;lA guinea pig skin test reportedly can detect the presence of
Legionella antigens when they are injected intradermally into immunized
guinea pigs. Biochemical tests for the presence of Legionella enzyme ac-
tivitigs are also under development.
/in ,u**ary, laboratory tests are available for early diagnosis and
retrospective confirmation of legionellosis. The most useful tests at this
time are the DFA and IFA tests in combination with culture on buffered
CYE agar directly or after passage through guinea pigs andlor chicken
embryo yolk sacs. The two fluorescent antibody tests require trained
laboratory personnel if test results are to be correctly interpreted.
Development of more selective culture media and diagnostic tests will
probably continue for some time. Therefore, it is necessary to keep
abreast of the literature and information available at state or local public
health laboratories. the Centers for Disease Control and current reviews. ;
rth"r, have not been controlled clinical trials, but the evidence
reported in the literature indicates that erythromycin and rifampin are
active against Legionella with in vitro assays, with in vivo models of
legionellosis, and in patients. Erythromycin gluceptate alone is given
intravenously in amounts ranging from 0.5 to 1.0 g every six hours. As
a patient's condition stabilizes, the dosage may be reduced and erythro-
mycin may be given orally. Antibiotic therapy is often continued for up
to three weeks. The combination of erythromycin and rifampin is recom-
mended if a patient is not responding to erythromycin alone. A con-
siderable amount of supportive therapy must be given to patients with
severe legionellosis.f
In vitro antimicrobial sensitivity studies with Legionella often pro-
duce results that are at variance with results of tests with animal models
and clinical experience. The aminoglycosides, penicillins,
cephalosporins, tetracycline, and chloramphenicol, although they may
be active in vitro, are not clinically effective. The ability of Legionella to
survive and multiply within macrophages may explain why antibiotics
that penetrate tissue are effective in vivo even though other antibiotics
that are active in vitro are not useful clinically. Also, most Legionella
produce a p-lactamase that is most active on the cephalosporin group of
antibiotics. Clearly, more controlled clinical trials will be needed before
questions concerning the efficacy of antibiotic therapy can be resolved.
a:,1:'Although a primary mode of transmission seems to be inhalation,
the source and nature of infective Legionella remains to be clearly
established. There is a tentative relationship between legionellosis cases
and aerosolized water. Therefore, preventive measures recommended at
this time are adequate disinfection of aerosol-producing equipment.
cooling towers and evaporative condensers heavily contaminated with
Legionella should be treated with effective levels of an appropriate
chemical biocide. Recirculating water systems may also be treated with
suitable ultraviolet light systems. Equipment heavily contaminated with
organic matter may not be effectively treated by chlorination, because the
organic material will neutralize the free chlorine. Therefore, free chlorine
levels must be monitored during disinfection procedures when calcium
hypochlorite is usedz'
One problem that prevents more effective preventative measures is
the lack of an efficient, cost-effective method to monitor the numbers of
viable Legionella in water samples that have a high background bacterial
flora. It is usually difficult to isolate Legionella from such samples by
direct plating on CYE agar, because the other bacteria in the sample fre-
quently overgrow or inhibit Legionella growth. The guinea pig in-
traperitoneal injection method is currently used, but it is time consuming
and pather expensive. A more efficient isolation method is clearly needed.
I Other preventive measures such as vaccination are being in-
vestigated. Immunization of guinea pigs has prevented subsequent infec-
tion, but more research will be required before the suitability of a vaccine
can be establishef
The author wishes to thank John J. Bozzola, PhD, Department of
Microbiology at The Medical College of Pennsylvania, Philadelphia, for
his contribution of electron micrographs to this chapter; critical reading
of the manuscript by Susan B. Dillon of the same department is gratefully
acknowledged. Portions of the work presented were supported in part by
The Whitaker Foundation and by The Biomedical Research Support
Grant Program of The Public Health Service, National Institutes of
Davis GS, Winn WC Jr, Gump DW, et al: Legionnaires'pneumonia after aerosol
exposure in guinea pigs and rats. Am Rev Respir Dis 1982;126:1050-1057.
England AC III, Fraser DW, Plikaytis BD, et al: Sporatic legionellosis in the
United States. The first thousand cases. Ann Intern Med 198l:94:.164-170.
Helms CM, Viner JP, Renner ED, et al: Legionnaires'disease among pneumonias
in Iowa (FY 1972-1978) II. Epidemiologic and clinical features of 30 sporatic
cases of L. pneumophila infection. Am J Med Sci l98l;281:2-13.
Hicklin MD, Thomason BM, Chandler FW, et al: Pathogenesis of acute Legion-
naires' disease pneumonia. Am J Clin Pathol 1980;73:480-487.
Jones GL, Hebert GA (ed): "Legionnaires." The Disease, the Bacterium and
Methodology. Atlanta, Georgia, Centers for Disease Control, 1979.
Katz S (ed): Legionellosis. Cleveland, Ohio CRC Press (in press).
Lattimer GL, Ormsbee RA, Legionnaires'Disease. New York, Marcel Dekker,
Inc, 1981.
Morris GK, Steigerwalt A, Feeley JC, et al: Legionella gormanii sp. nov. J. CIin
Microbio I 1980;12:7 18-721.
Orrison LH, Cherry, WB, Tyndall RL, et al: Legionella oakridgensls: Unusual
new species isolated from cooling tower water. Appl Environ Microbiol
Reingold AL, Band JD, Legionellosis, in Easmon CSF, Jaszewicz J (eds):
Medical Microbiology, vol l, New York, Academic Press, 1982.
ResearchGate has not been able to resolve any citations for this publication.
We describe a new species of Legionella represented by 10 strains isolated from industrial cooling towers. Legionella oakridgensis differed genetically from the other seven species of Legionella in DNA hybridization studies and differed serologically in direct fluorescent-antibody tests. The new species, unlike all other species except L. jordanis, did not require added L-cysteine for growth in serial transfer on charcoal-yeast extract agar. L. oakridgensis, as well as three other species tested, required L-cysteine for primary isolation from animal tissues. L. oakridgensis was the only species of Legionella that failed to produce alkaline phosphatase at pH 8.5. In all other respects, it resembled other species of Legionella, including having a high content of branched-chain cellular fatty acids and being pathogenic for guinea pigs. These bacteria have not yet been associated with human disease, but they are potential causes of legionellosis.
We developed an animal model of Legionnaires' pneumonia to permit study of aerosol infection, pathogenesis, and pulmonary host defense mechanisms in this disease. Guinea pigs and rats were exposed in a nose-only inhalation facility for 30 min to an aerosol of Burlington serogroup 1 Legionella pneumophila. Lungs contained 10(3) to 10(4) L. pneumophila immediately after exposure. Both guinea pigs and rats developed pneumonia, with 100% infectivity by microbiologic, histologic, and serologic criteria. Guinea pigs demonstrated illness, fever, and 56% mortality; rats showed little illness and 11% mortality. In both species, diffuse patchy pneumonitis coalesced and consolidated as the disease progressed. Aerosol challenge with 3H-L. pneumophila showed exponential growth of the bacteria in the lungs of both species. Guinea pigs and rats can be infected by aerosol exposure to L. pneumophila to produce a disease that closely simulates human Legionnaires' pneumonia. Rapid initial intrapulmonary growth suggests that resident lung defense mechanisms are quite ineffective against L. pneumophila, and that recruited or immunospecific defenses may be more critical in the outcome of infection. The difference in severity of illness between guinea pigs and rats may be exploited for different experimental designs.
Immunofluorescent study of the lungs in cases of fatal suspected acute Legionnaires' disease enabled confirmation of the presence of Legionella pneumophila. In addition, probable pathogenetic mechanisms that had not been as clearly visualized by light microscopy became apparent: the retrograde involvement of the larger bronchioles and proximal airways, invasion of the interstitium, extension to pleura, and lymphatic and hematogenous spread. Organisms were demonstrated to occur in the liver and spleen of one patient. The development of technics for the earliest possible diagnostic verification of Legionnaires' disease, with specimens obtained as untraumatically as practical from selected sites and screened by specific immunofluorescent microscopic examination, should contribute to patient survival.
A new species of Legionella was isolated from soil collected from a creek bank. The name Legionella gormanii sp. nov. is proposed.
As of 30 September 1979, 1005 confirmed cases of sporadic legionellosis caused by Legionella pneumophila serogroups 1 to 4 in U.S. residents had been reported to the Centers for Disease Control; 19% were fatal. All but 2% of the 1005 cases were associated with pneumonia documented by chest radiograph. About 75% of the cases occurred in June through October. The risk of acquiring sporadic legionellosis was increased among males and persons 50 years or older; persons with renal disease necessitating dialysis or transplantation, with chronic bronchitis or emphysema, with diabetes mellitus, and with cancer (10 selected sites or types); persons who smoke; and persons being treated with immunosuppressive drugs. Increasing age and chronic bronchitis or emphysema were associated with increased risk of death. The sensitivity of culturing L. pneumophila from specimens positive by direct immunofluorescence was estimated to be 45%. The distribution of serogroups 1, 2, 3, and 4 of L. pneumophila in 57 fresh, not previously examined direct fluorescent antibody-positive specimens was 84%, 11%, 4%, and 2%, respectively; all 26 strains isolated from these specimens were of one of these four serogroups.
We reviewed retrospectively the clinical records of 30 cases of sporadic Legionella pneumophila infection that occurred in Iowa between FY 1972 and 1978. Cases occurred throughout the year, most between May and December. Twenty-one male patients and 9 female patients ranging in age from 5-80 years were infected. Half the patients smoked or had an underlying illness; five were receiving corticosteroids or immunosuppressive therapy. Occupations and exposures related to hospitals, construction and travel were common; four patients had been exposed to birds. In addition to L. pneumophila infection, six patients had evidence of infection with a viral, mycoplasmal, bacterial, mycobacterial or fungal pathogen; three had had preceding dental infections. Twenty-seven cases were pneumonias visible on radiographs. Fever, cough, chills, myalgia and rales occurred inover half the cases. Headache, gastrointestinal symptoms and encephalopathy also were seen. Upper respiratory symptoms were uncommon. Urinalysis and blood studies often suggested renal and hepatic involvement, but other routine laboratory diagnostic tests were not helpful. All but two patients were hospitalized; seven required intensive care. The median duration of hospitalization was 12 days. Two patients who did not receive erythromycin or tetracycline therapy died.
The Disease, the Bacterium and Methodology
  • G L Jones
  • G A Hebert
Jones GL, Hebert GA (ed): "Legionnaires." The Disease, the Bacterium and Methodology. Atlanta, Georgia, Centers for Disease Control, 1979.