Scientific

Bacterial infection of the lower respiratory tract in 34

horses
DJ RACKLYEFTa and DN LOVEb

Objective To investigate associations between the bacteri- diseased lungs to organisms identified as normal oral flora is
ology and aspects of history, clinical presentation, outcome also now well established.1-3 Biberstein et al4 first reported the
and pathology of lower respiratory tract disease of 34 horses. isolation of Bacteroides melaninogenicus (possibly Prevotella
Procedure Detailed aerobic and anaerobic bacteriological melaninogenica) from equine respiratory tract samples.
investigations were performed on clinical specimens from Subsequent studies have reported the most frequently isolated
horses with pneumonia, lung abscessation and necrotic pneu- anaerobic bacteria from equine pleuropneumonia as Bacteroides
monia with or without pleurisy in an attempt to identify those spp, including B fragilis, Peptostreptococcus anaerobius and other
bacteria that might contribute to the initiation and progression Gram-positive cocci and Fusobacterium spp.2,3,5-10
of infection. In one study of 46 cases of equine pleuropneumonia where
Results Bacteria were cultured from 33 of the 34 horses. In anaerobes were isolated from 21 horses, a significant association
ten cases, only aerobic/facultatively anaerobic isolates were was found between the isolation of anaerobic bacteria and death
cultured while aerobic/facultatively anaerobic bacteria and or euthanasia.7 Therefore, recognition of an anaerobic infection
obligately anaerobic bacteria were isolated in the other 23 may influence the prognosis. Clinical findings that are indica-
cases. Moderate to large numbers of anaerobic bacteria were tive of anaerobic infection are a putrid odour of the breath or
isolated only when the estimated duration of illness was at pleural fluid samples and the presence of gas echoes within the
least five days. Bacteria were not cultured from 12 of the pleural cavity or lung abscesses detected by ultrasonography.7,11
pleural fluid samples but were always cultured from With the temporal association between certain events such as
pulmonary samples (either transtracheal aspirates from live transportation and strenuous exercise and the onset of equine
horses or pulmonary lesions at necropsy). Streptococcus equi
pneumonia, an accurate history may contribute to the diagnosis
subsp zooepidemicus was isolated in the three cases where
only one bacterial species was cultured. In the other 30 cases, of such infections in horses. 12
multiple species were isolated. These included most often and In this study we performed detailed bacteriological investiga-
in greatest numbers, Streptococcus equi subsp zoo- tions on clinical specimens from horses with pneumonia, with
epidemicus, Pasteurellaceae , Escherichia coli , anaerobic or without pleurisy, in an attempt to identify those bacteria that
cocci, Eubacterium fossor, Bacteroides tectum , Prevotella might contribute to the initiation and progression of infection.
heparinolytica, Fusobacterium spp, and pigmented members We also investigated whether there was any association between
of the genera Prevotella and Porphyromonas. Aerobic/faculta- the bacteriology of such cases and aspects of history, clinical
tively anaerobic organisms were isolated from 97% of horses, presentation, outcome and pathology.
while obligately anaerobic organisms were cultured from 68%
of horses.
Materials and methods
Conclusion There was no association between the isola- Selection of cases
tion of any specific bacterium and the outcome of disease.
Horses were included if they had been diagnosed clinically as
However, obligately anaerobic bacteria (such as anaerobic
cocci, Bacteroides tectum, P heparinolytica and
having pneumonia with or without pleurisy with the diagnosis
Fusobacterium spp) and the facultatively anaerobic species supported by appropriately obtained samples from the lower
Escherichia coli, were recovered more commonly from horses respiratory tract in the live animal and/or by pathological
that died or were euthanased than from those that survived. changes observed at necropsy. Horses were included only if
There was an association between failure of horses to recover samples had been collected appropriately and preserved for
from pleuropneumonia and delay in diagnosis and initiation of aerobic and anaerobic bacteriological processing.
treatment.
Aust. Vet J 2000;78:549-559 History and physical examination
Key Words: Horse, equine pleuropneumonia, lower respiratory tract, anaerobic
The historical information obtained from owners or
bacteria
attending veterinarians included: age, breed and sex of horse;
whether there had been any ‘stressful’ event prior to the onset of

E
quine pneumonia, which may progress to lung abscessa- the current respiratory illness; whether the horse had received
tion or necrotic pneumonia with or without pleuritis, is antimicrobial treatment; the duration of the current illness; and
most commonly reported to be associated with multiple the time at which treatment was instituted relative to the onset
bacterial species, with the b-haemolytic streptococci (nearly of illness (Table 1). In addition, the following details of clinical
always Streptococcus equi subsp zooepidemicus), Pasteurella spp examination were provided either by the attending veterinarian
and Escherichia coli the most commonly isolated facultative
species. The importance of anaerobic bacteria in infections of
the lung and pleural space, and the resemblance of isolates from
IM Intramuscular
IV Intravenous
aSatur Veterinary Clinic, Bunnan Road, Scone, New South Wales 2337 PO Per os
bDepartment of Veterinary Anatomy and Pathology, The University of Sydney,
TTA Transtracheal aspirate
New South Wales 2006

Aust Vet J Vol 78, No 8, August 2000 549

Scientific

Table 1. Details of signalment, history, samples collected and bacterial growth and outcome for 34 horses with pneumonia, with or without pleuritis.

Horse Age Event Onset of Outcomeb Time of Sample Aerobic/FA growthc Anaerobic growthc
treatmenta sampling
(d) relative to onset
of illness (d)

51 6y transport <1 paddock-soundd <1 TTA +++ -

52 4y transport <1 paddock-sound <1 TTA +++ ++

54 4y transport <1 paddock-sound <1 TTA ++ -

56 18 m transport <1 raced <1 TTA +++ +

60 2y strenuous <1 raced 3e TTA +++ -

exercise pleural fluid(R) - -

67 1y GA 4 euthanased 5e TTA ++++ ++++

pleural fluid(R) +++ -
pleural fluid(L) ++ -
20e abscess(N) ++++ +++

68 2y strenuous 7 euthanased 15e TTA +++ (+)

exercise pleural fluid(R) (+) -
pleural fluid(L) - -

75 2y strenuous <1 raced 4e pleural fluid(R) - -

exercise pleural fluid(L) - -

78 4y strenuous <1 raced 7e pleural fluid(R) ++ -

exercise pleural fluid(L) + -

79 6y strenuous 21 euthanased 21e TTA ++++ ++++

exercise pleural fluid(R) ++++ ++++
pleural fluid(L) ++++ ++++
35e pleural fluid(R) (+) -
pleural fluid(L) + -
40e pleural fluid(N) ++++ ++++

88 2y intratracheal <1 euthanased 7e TTA ++++ +

drench (laminitis) 14e pleural fluid(R) (+) -
40e abscess (N) ++++ -

105 2y strenuous <1 raced 5e TTA +++ -

exercise pleural fluid(R) - -

pleural fluid(L) - -

106 1y transport <1 raced 2e TTA ++++ (+)

pleural fluid(R) ++++ +
pleural fluid(L) ++++ -

107 2y strenuous 4 euthanased 11e TTA ++++ ++++

exercise (acute diarrhoea) pleural fluid(R) ++ (+)

pleural fluid(L) +++ +

21e pleural fluid(R) (N) ++++ -

abscess (N) ++++ ++++

121 4y transport <1 euthanased 16e abscess (N) +++ +++

(laminitis)

124 1y transport <1 paddock-sound 21e TTA +++ +++

pleural fluid(R) ++ +++
pleural fluid(L) + +++

132 2y strenuous 2 euthanased 25e TTA +++ +++

exercise (laminitis) pleural fluid(R) +++ +
pleural fluid(L) (+) -
abscess(N) +++ +

158 2y strenuous <1 raced 14e TTA ++ +

exercise

164 2y transport and 7 died 10e abscess(N) ++++ ++++

strenuous abscess(N) +++ ++++
exercise pleural fluid (R) (N) ++++ ++++

550 Aust Vet J Vol 78, No 8, August 2000

 Scientific

Table 1. (Continued)

Horse Age Event Onset of Outcomeb Time of Sample Aerobic/FA growthc Anaerobic growthc
treatmenta sampling
(d) relative to onset
of illness (d)

171 2y URT “viral” 7 euthanased 14e TTA +++ +++

infection abscess (N) +++ +++
pleural fluid(R) (N) ++ +++

175 1y diarrhoea 3 euthanased 9e TTA ++++ ++

abscess (N) +++ ++

176 2y half-drowning 3 euthanased 5e pleural fluid (R) ++ -

pleural fluid (N) ++ ++

191 3y strenuous <1 raced < 1e pleural fluid (R) + -

exercise pleural fluid (L) (+) -

208 1y transport <1 euthanased 7e abscess (N) ++ ++

(laminitis) pleural fluid (R) ++ (+)
pleural fluid (L) ++ (+)

210 1y unknown <1 paddock-sound 5e TTA ++++ +

214 15 m transport <1 euthanased < 1e TTA ++++ -

(acute diarrhoea) pleural fluid(R) + -
pleural fluid(L) - -

25e abscess (N) +++ -

242 3y strenuous 2 died 7e necrotic lung (N) +++ +++

exercise

243 2y strenuous <1 euthanased 7e TTA +++ +++

exercise pleural fluid R) + -
pleural fluid(L) - -
21e TTA +++ +++
pleural fluid (R) (N) +++ ++++

244 8m transport <1 paddock-sound 3e TTA +++ +++

245 7y URT “viral” <1 raced 5e TTA +++ -

infection

246 2y strenuous 2 euthanased 14e TTA +++ -

exercise pleural fluid(R) - -

259 2y strenuous 4 euthanased (laminitis, 10e necrotic lung +++ +

exercise acute diarrhoea)
and EPIH

260 9m debility 7 euthanased 1e TTA +++ -

pleural fluid (R) ++++ -
7e pleural fluid(L) +++ -
pleural fluid (N) - -

261 2y strenuous <1 died (acute diarrhoea) 10e TTA +++ ++

exercise pleural fluid (R) ++++ ++
14e pleural fluid (L) +++ ++
abscess (N) +++ ++
pleural fluid (R) - -
pleural fluid (L) - -

aRefers to interval between onset of clinical signs and the commencement of treatment
bCondition in parentheses represents non-respiratory complication(s) which lead to death or euthanasia
c- = no growth on culture, no organisms visible on direct smears; (+)= isolated indirectly via broth only, no organisms seen on smear; +=1-10 colonies on direct plate,
occasional organism visible on smear; ++= 11-50 colonies on direct plates, small numbers of organisms visible on smear; +++= 51-200 colonies on plates, moderate
numbers of organisms visible on smear; ++++= over 200 colonies (or usually innumerable) on plates, many organisms visible on smear
dSufficiently sound to graze at pasture; unfit for work
eHorse received antimicrobial therapy prior to, or at the time of sampling

M=stallion or colt; F=mare or filly; G=gelding; TB=Thoroughbred; SB=Standardbred; XB=crossbred; TTA= transtracheal aspirate; EPIH=exercise-induced pulmonary
haemorrhage; URT=upper respiratory tract; R=right; L=left; N=necropsy

Aust Vet J Vol 78, No 8, August 2000 551

Scientific

or were assessed at the time of examination: the presence of 5% sheep blood and zone sizes interpreted according to criteria
depression, loss of appetite, respiratory rate (elevated over 15 supplied by the manufacturer.
breaths per min), fever (over 38.5oC), abnormal thoracic Antibiotic susceptibility of anaerobic isolates was tested and
auscultation, exudate in the trachea (observed by endoscopy or interpreted using the broth disc method against penicillin (2
obtained by TTA), a cough, nasal discharge, pleuritis (fluid U/mL), amoxycillin (4 mg/mL), doxycycline (6 mg/mL),
detected by radiography, ultrasonography or thoracocentesis erythromycin (3 mg/mL), and metronidazole (2 mg/mL).20
with cytological confirmation that the pleural fluid was an
inflammatory exudate) and presence of a foetid odour to the Results
horse’s breath or in pleural fluid samples. Clinical cases
Over a 30-month period, 87 satisfactory samples were obtained
Necropsy from 34 horses. The details of signalment and history are pre-
Necropsy was performed in most cases and gross pathology sented in Table 1. There were 21 females, eight entire males and
and the distribution of lesions were documented. five geldings. Thirty two horses were Thoroughbreds, with one
Standardbred and one crossbred horse. A prior ‘stressful’ event
Collection of samples from the lower respiratory tract
associated with the onset of disease was identified in 32 cases. At
With the exception of horses 105, 124 and 160 (each with a
the time(s) of sampling, most horses were receiving or had
TTA and fluids obtained by thoracocentesis by referring veteri-
received antimicrobial treatment. Only seven samples (from five
narians), samples were collected by the senior author.
horses) were collected before antimicrobial therapy had com-
Respiratory tract samples from living horses were collected by
menced. In many cases treatment had been initiated within 24
TTA.13 Thoracocentesis was performed in the right hemithorax
h of the onset of illness (20 horses), but in other cases recogni-
at the level of the point of the elbow in the seventh intercostal
tion of pleuropneumonia (and therefore treatment) had been
space. In the left hemithorax, the procedure was performed 4 to
delayed (up to 21 days in horse 79). Serial samples were
6 cm above the point of the elbow in the eighth or ninth inter-
obtained from eight horses. The remaining 26 horses were sam-
costal space. In some cases the site for thoracocentesis was deter-
pled on a single occasion.
mined by ultrasonography. Fluid was aspirated into a syringe.
Samples of lower airway exudate collected at necropsy were Physical examination
obtained by making an incision with a sterile scalpel blade in a The severity of clinical signs varied greatly among the 34
bronchus through a seared area and directly aspirating material cases. In early cases clinical signs were often not obvious. The
through a ten gauge catheter. In the few cases where there was earliest cases examined (horses 51, 52 and 54) had abnormali-
only a small amount of exudate present or if it was very tena- ties immediately on arrival following air transportation from
cious, 5 to 10 mL of PBS was injected and reaspirated. England. In one case (horse 52) the only abnormality detected
Contents of abscesses were collected by searing the outer was a gurgling sound of exudate in the trachea. Another two
capsule and aspirating the contents through a 14 gauge needle. cases examined at a similarly early stage (horses 51 and 54) were
In cases where there was only a small amount of pus or if it was also febrile and were depressed and not eating. Horses examined
caseous, a sample was collected by incising the seared abscess up to 72 h following a ‘stressful’ event, but within 24 h of
capsule with a sterile scalpel blade and sampling the contents exhibiting clinical abnormality, showed fever, depression and
with a bacteriological swab moistened with sterile physiological inappetence, increased respiratory rate, varying amounts of
saline. Pleural fluid samples were collected directly from the tracheal exudate, and harsh respiratory sounds dorsally and
thoracic cavity as soon as the rib cage had been reflected. dullness ventrally on auscultation (horses 56, 60, 75, 106, 191
To optimise survival of obligately anaerobic bacteria, syringes and 214). Some cases (horses 56 and 191) were also reluctant to
containing samples had all air expelled from capped syringes move with pleurodynia and small amounts of pleural effusion
and bacteriological swabs were stored in anaerobic specimen (up to 2 L obtained). In more advanced cases, estimated to have
collectors. been ill for at least 5 d (Table 1), auscultation revealed more
extensive changes and, where performed, radiography demon-
Processing of samples
strated ventral thoracic density and ultrasonography showed
Two smears were made as soon as possible following sample
pleural effusion (up to 20 L in horse 208), often with extensive
collection. One was stained with DiffQuik stain for differential
fibrin deposition. Ten horses had a putrid odour to their breath
cell counts and the other was examined for bacteria using
and/or pleural fluid samples (horses 67, 79, 107, 121, 124, 164,
Burke’s Gram stain. Samples were processed for bacterial isola-
175, 242, 243 and 261) and all had shown clinical signs for 5 d
tion on arrival at the laboratory (delay interval ranged from 10
or more when first examined. Weight loss was also a feature of
min to 24 h although most were processed within 2 h of collec-
more advanced cases.
tion). In addition to direct plating onto sheep blood agar plates,
Nasal discharge and coughing were inconsistent findings.
which were incubated aerobically and anaerobically, samples
Nasal discharge was present in 13 horses (horses 56, 67, 88,
were also inoculated into broth media (cooked meat broth for
107, 124, 164, 171, 176, 214, 242, 243, 246 and 259) and was
subsequent aerobic subculture and pre-reduced brain heart
usually mucopurulent, but was also haemorrhagic in five cases
infusion broth for anaerobic subculture).
(horses 67, 88, 107, 242 and 259). In 11 horses, the only
Phenotypic characterisation of bacteria was carried out
prominent sign was coughing (horses 51, 107, 124, 132, 164,
according to methods and criteria described previously.14-18
171, 175, 214, 243, 244 and 261).
Antimicrobial susceptibility testing In 21 horses (horses 60, 67, 78, 79, 88, 106, 107, 121, 124,
The antimicrobial susceptibility of aerobic and/or faculta- 132, 164, 171, 175, 176, 191, 208, 214, 242, 243, 260 and
tively anaerobic bacteria was tested by the Kirby-Bauer disk 261), pneumonia was accompanied by a septic pleural exudate.
diffusion method 19 on Oxoid Isosensitest agar with or without A nonseptic pleural exudate was present in a further five cases

552 Aust Vet J Vol 78, No 8, August 2000

 Scientific

(horses 68, 75, 105, 246 and 259). In one case (horse 56) there cm diameter) where pulmonary parenchyma was obliterated
was evidence of pleurodynia, but thoracocentesis was unsuc- and replaced by liquified, foul-smelling brown-grey material
cessful. In another case (horse 158) pleural fluid was visualised traversed by trabecular remnants of pulmonary vasculature and
ultrasonographically but was not obtained by thoracocentesis. connective tissue.
Six horses (horses 51, 52, 54, 210, 244 and 245) had no clinical In other cases there was pulmonary abscessation. Some
evidence of pleuritis and pleural fluid samples were either not abscesses were large and filled with grey, foul-smelling pus,
obtained or were cytologically normal. while others were multiple, small, caseous and surrounded by a
thick fibrous capsule. If there was a large abscess or necrotic
Treatment, clinical course and outcome cavity in one lobe, there were often miliary abscesses in the
All horses received treatment with antimicrobial agents and ventral part of the contralateral lung lobe.
the duration of treatment ranged from 3 to 56 days. Drug and In some cases, areas of cavitation communicated with at least
dosage regimens varied greatly among individual cases. Procaine one airway, but there was rarely free drainage of large amounts
penicillin was the agent most commonly used, administered of pus into the trachea. Abscesses or necrotic areas sometimes
IM, often in combination with sodium benzyl penicillin IV, had fistulated into the pleural cavity, often on the mediastinal
neomycin sulphate (IV or IM), gentamicin sulphate (IV or IM) aspect of the affected lobe, to form abscesses (horses 79, 107,
and/or metronidazole PO. Trimethoprim and sulphonamide 121 and 243).
combinations (IV or PO), oxytetracycline hydrochloride (IV) Most horses exhibited extensive fibrinous pleuritis, either
and ampicillin trihydrate (IM) were also used. Drainage of bilateral or unilateral. Both visceral and parietal pleural surfaces
pleural effusion was also performed in most cases by placement were covered with a thick layer (up to 1 cm) of fibrin, frequently
of thoracic drainage tubes. Supportive therapies included the with extensive loculation and adhesions. Accompanying this
administration of non-steroidal anti-inflammatory agents were varying amounts of pleural exudate which ranged from
(phenylbutazone and/or flunixin meglumine), IV fluid therapy, golden and turbid to foul-smelling, brown and fibrinous. The
and mucolytic or bronchodilator therapy. In some cases amount and character of the pleural exudate was often different
attempts were made to administer antibiotics directly by nebuli- on the left and right sides of the pleural cavity. In one case
sation into the respiratory tract. (horse 88) pleuritis had resolved, and established fibrous adhe-
Fifteen horses survived and of these, nine returned to racing sions were present between the affected lung lobe, pericardium
(Table 1). In all 15 horses, treatment had commenced within 24 and thoracic wall.
h of onset of clinical signs. In some cases (horses 67, 88, 132, 214 and 261) pulmonary
Nineteen horses died or were euthanased. In 11 cases this was lesions (pneumonia, necrosis or abscessation) predominated,
as a result of their respiratory disease only but in eight cases and in others (horses 68, 79 and 121) thoracic empyema was
complications of pleuroneumonia and/or its treatment also the predominant lesion. In three cases (horses 88, 121 and 243)
contributed to their failure to recover. Where treatment had pulmonary lesions were exclusively in the right lung. In ten
been instituted after 24 h, all 13 horses (68%) died or were further cases (horses 67, 79, 107, 132, 164, 175, 214, 259, 260
euthanased. In this group, clinical signs had been evident at and 261) pulmonary lesions occurred bilaterally, but were more
least 2 d (range 2 d to 21 d). In the remaining six horses which severe on the right side. In only one case (horse 171) were
died, treatment had been commenced within 24 h of onset of lesions present only in the left lung. In one case pneumonitis
clinical signs. and fibrinous pleuritis were equally severe on both sides of the
Nine of the 10 horses, which had a putrid odour on their thorax (horse 68). Where there was pulmonary abscessation
breath and/or pleural fluid samples, died or were euthanased and/or necrosis (14 cases) 13 occurred in the right and only one
because of their pulmonary disease or complications associated in the left lung.
with treatment. All nine horses had disease of 7 d or longer
(range 7 to 40 d) at time of death. Cytology
Differential cell counts from all pleural fluid, TTA and
Gross pathology necropsy samples demonstrated over 50% neutrophils. In the
Necropsies were performed on 15 of the 19 horses that died most severe cases these were undergoing severe degeneration
or were euthanased. The estimated duration of illness ranged and, in samples from abscesses or areas of necrosis, cell
from 8 to 80 d. Details of pathological findings in individual morphology was almost unrecognisable. In most cases Gram-
cases are presented in Figure 1. In all but one case (horse 259), stained smears demonstrated bacteria (ranging from less than
pulmonary lesions were cranioventral in distribution, concen- one organism to more than 100 organisms per high power
trated in those areas of the lung that were supplied by the first field). There were often many different bacterial morphotypes
bronchi to branch ventrally. In horse 259, pneumonia was asso- present. Although protein and total leukocyte counts were not
ciated with pulmonary haemorrhage in the caudal lobes of the determined, the samples were considered to be inflammatory,
lung. based on the presence of a large proportion of neutrophils. In
In all horses there were areas where lung parenchyma was most cases bacteria were observed on smears and/or cultured.
non-aerated, firm, red and moist, and in those cases where there
was significant pleural effusion, the lung parenchyma was Bacteriological findings and outcomes
collapsed up to the dorsal extent of pleural fluid. In some cases Eighty seven samples from 34 horses were cultured aerobi-
there was very little aerated pulmonary parenchyma except in cally and anaerobically. A total of 435 isolates were obtained.
the most dorsal parts of the lungs. However ten isolates either failed to grow following the initial
In some cases there was pulmonary parenchymal necrosis, description and subculture from anaerobic plates (two isolates)
either consisting of multiple small areas of friable material or grew initially but subsequently failed to grow or did not
surrounded by a white capsule or very large cavities (up to 20 correspond to original descriptions (eight isolates). Where

Aust Vet J Vol 78, No 8, August 2000 553

Scientific

Figure 1. Schematic representation of distribution of lesions in necropsied horses showing the duration of disease and extent of the
lesion. Figure at bottom of each diagram is the horse number and that in brackets represents the estimated duration of illness at time of
necropsy. The areas within the dashes represent mediastinal abscessation.

554 Aust Vet J Vol 78, No 8, August 2000

 Scientific

Table 2. Summary of bacteriological results from lower respiratory infection in 34 horses

Aerobic and/or FA isolates (n=148) No. of horses from Obligately anaerobic isolates (n=144) No. of horses from
which isolate was which isolate was
cultured (%) cultured (%)

Aerobic/FA G+ cocci (72) 31 (91%) Anaerobic G+ cocci (35) 15 (44%)

Micrococcus spp (1) 1 (3%) Peptostreptococcus anaerobius (9) 9 (26%)
Other G+ aerobic cocci (1) 1 (3%) Other anaerobic G+ cocci (26)2,9,24 14 (41%)
Coagulase-positive Staphylococcus spp (4) 3 (9%)
Coagulase-negative Staphylococcus spp (4) 4 (12%) Anaerobic G+ bacilli (44) 19 (56%)
Other G+ FA catalase-positive cocci (1) 1 (3%) Clostridium spp (9) 8 (24%)
Streptococcus equi subsp zooepidemicus (28) 28 (82%) Eubacterium fossor (11)9,14 11 (32%)
Enterococcal Streptococcus spp (2) 2 (6%) Propionibacterium spp or Arachnia sp (1) 1 (3%)
Group D nonenterococcal Streptococcus spp (1) 1 (3%) Other anaerobic G+ nonsporing bacilli (23) 9 (26%)
Viridans Streptococcus spp (30) 14 (41%)
Anaerobic G- cocci (4) 4 (12%)
Aerobic/FA G+ bacilli (12) 8 (24%) Veillonella spp (2) 2 (6%)
Bacillus spp (2) 2 (6%) Megasphaera sp (1) 1 (3%)
Aerobic G+ nonsporing bacilli (2) 2 (6%) Other anaerobic G+ cocci (1) 1 (3%)
Actinomyces spp (8) 5 (15%)
Anaerobic G- bacilli (61) 18 (53%)
Aerobic/FA G- bacilli (64) 29 (85%) Bacteroides spp
Acinetobacter spp (5) 5 (15%) B tectum (8) 2,9,16 8 (24%)
Pseudomonas aeruginosa (5) 5 (15%) B fragilis (4) 4 (12%)
Other Pseudomonas spp (1) 1 (3%) P heparinolytica (7) 7 (21%)
Escherichia coli (14) 14 (41%) P zoogleoformans (2)15,25,26 2 (6%)
Proteus sp (1) 1 (3%) pigmented Porphyromonas/Prevotella spp (7)27 7 (21%)
Salmonella Typhimurium (1) 1 (3%) nonpigmented Prevotella spp (26) 14 (41%)
Other G- FA, oxidase-negative bacilli (8) 7 (21%) Fusobacterium spp (7) 7 (21%)
Pasteurellaceae (27)21-23 22 (65%)
Aeromonas sp (1) 1 (3%)
Other G- FA, oxidase-positive, motile bacilli (2) 2 (6%)

Total number of horses from which aerobic/FA bacteria 33 (97%) Total number of horses from which obligately 21 (68%)
were cultured anaerobic bacteria were cultured

FA = facultatively anaerobic; G+ = Gram positive; G- = Gram negative

Figures in brackets are numbers of isolates in each category ; bolded figures represent total for each category

isolates with the same characteristics were cultured from Table 3. Species cultured in large numbersa from lower respiratory infec-
multiple samples from the same horse, only one isolate from tion in 34 horses

each horse was reported. Of the remaining isolates that were Identity of isolate No. of cases from which
species were cultured in
characterised, 144, 132 and 16 were obligately anaerobic, facul- large numbers (%)
tatively anaerobic and obligately aerobic bacteria respectively. A
Streptococcus equi subsp zooepidemicus 18 (53%)
full description of these isolates can be found elsewhere.17
Pasteurellaceae 8 (24%)
In only one case (horse 75), in which pleural fluid was the
Pigmented Porphyromonas/Prevotella spp 6 (18%)
only sample obtained, were no bacteria cultured and/or seen on Escherichia coli 5 (15%)
direct smears. In 10 cases, aerobic/facultatively anaerobic Bacteroides tectum 3 (9%)
isolates only were cultured (Table 1) whereas aerobic/faculta- Nonpigmented Prevotella spp 3 (9%)
tively anaerobic bacteria were isolated together with obligately Viridans Streptococcus spp 2 (6%)
anerobic bacteria in the other 23 cases. Moderate to large Pseudomonas aeruginosa 2 (6%)
numbers of anaerobic bacteria were isolated only where the esti- Obligately anaerobic Gram positive cocci 2 (6%)
mated duration of illness was at least 5 d (Table 1). (not Peptostreptococcus anaerobius)
Bacteria were not cultured from 12 (26%) of the pleural fluid Other obligately anaerobic Gram positive bacilli 2 (6%)
samples. In another nine samples, a positive culture was Fusobacterium spp 1 (3%)
obtained via enrichment broth only. However, bacteria were Bacteroides fragilis 1 (3%)
always cultured from pulmonary samples (either TTA from live Veillonella spp 1 (3%)
horses or pulmonary lesions at necropsy) and in all but two of Eubacterium fossor 1 (3%)

such samples moderate to large numbers of bacteria were Acinetobacter sp 1 (3%)

isolated. In those horses from which bacteria were isolated from Actinomyces sp 1 (3%)
Coagulase-positive Staphylococcus sp 1 (3%)
both pleural fluid and pulmonary samples, additional bacterial
isolates could be isolated frequently from the latter. Most a = +++ or ++++ (see Table 1 footnote for explanation)

Aust Vet J Vol 78, No 8, August 2000 555

Scientific

Table 4. Antimicrobial susceptibility of aerobic and/or facultatively anaerobic bacteria recovered from lower respiratory infection in 34 horses

Identity of isolate Proportion (%) of isolates susceptible to antimicrobial agentsa

(No. tested)
Penicillin Tetracycline Neomycin Gentamicin Trimethoprim Sulfamethoxazole TMS/SMZ Other agents

Equivalent MICb £ 0.12 mg/mL £ 4 mg/mL - £ 6 mg/mL £ 4 mg/mL £ 100 mg/mL £ 2/38 mg/mL

Micrococcus sp (1) R S S S R Erythromycin (S)

Other aerobic G+ cocci (1) S R S S R Ampicillin (S), chloramphenicol (R)
Coagulase-positive 0 75 100 Ampicillin (0), chloramphenicol (100),
Staphylococcus spp (4) cloxacillin (100), erythromycin (100),
lincomycin (50)
Coagulase-negative
Staphylococcus spp (4) 0 25 50 50 0 Ampicillin (0), chloramphenicol (50),
cloxacillin (25), erythromycin (25),
lincomycin (0)
Bacillus sp (1) R S S S Ampicillin (R)
Actinomyces spp (3) S Amoxycillin (100), doxycycline (66),
erythromycin (100), metronidazole (33)
Acinetobacter spp (5) 0 40 100 100 0 Ampicillin (20)
Pseudomonas aeruginosa (3) 0 0 0 100 0 Ampicillin (0), chloramphenicol (0)
Escherichia coli (14) 0 36 29 100 21 21 21 Ampicillin (14), amoxycillin/clavulanic
acid (71), cefoxitin (100),
cephaloridine (0)
Proteus sp (1) R R S S Ampicillin (S), chloramphenicol (R)
Salmonella Typhimurium (1) R R S S S R S Ampicillin (S), amoxycillin/clavulanic
acid (S), cefoxitin (S),
cephaloridine (S), chloramphenicol (S)
Other FA G- oxidase - 0 29 33 56 60 60 60 Ampicillin (14), amoxycillin/clavulanic
bacilli (7) acid (43), cefoxitin (57),
cephaloridine (0)
Aeromonas sp (1) R R R S R Ampicillin (R)
Other FA G- oxidase +
motile bacilli (1) S S R S S Chloramphenicol (S)

Pasteurellaceae (22) 32 86 45 86 68 18 68
aWhere results are given for only one isolate S = sensitive, R = resistant; where results are given for multiple isolates they are given as % of isolates sensitive to the
antibiotic tested
bMinimum inhibitory concentration data for human isolates according to M2-A3 Performance Standards for antimicrobial disc susceptibility tests (Third Edition). Approved
Standard December 1984. National Committee for Clinical Laboratory Standards, Villanova, Philadelphia.
TMS = Trimethoprim; SMZ = Sulfamethoxazole; FA = facultatively anaerobic; G+ = Gram positive; G- = Gram negative

notably, in five horses, anaerobic bacteria were cultured from (Table 1). Although a putrid odour could not be detected in the
pulmonary samples, but not from pleural fluid samples. remaining 24 horses, anaerobes were isolated nevertheless, from
The total number of different species characterised per horse 13. Seven of these horses died.
ranged from one to 17 with a mean of seven. In only three cases Of the 14 horses from which E coli was isolated as part of the
(horses 78, 191, 145) was only one species recovered (S equi aerobic/facultatively anaerobic flora (horses 56, 79, 88, 106,
subsp zooepidemicus). In the other 30 cases from which bacteria 107, 121, 132, 158, 164, 175, 208, 243, 259, 261), 11 died or
were recovered, infection was polymicrobial. Table 2 were euthanased as a consequence of untoward sequelae of treat-
summarises the bacteriological results according to the identity ment.
of isolates and their overall prevalence. In three cases (horses 51,
54, 260) where only two species were characterised, Antimicrobial susceptibility of clinical isolates
Pasteurellaceae accompanied S equi subsp zooepidemicus. These Disc diffusion susceptibility results for 69 aerobic and/or
two species of bacteria were also isolated most commonly in facultatively anaerobic isolates are presented in Table 4.
large numbers from diseased horses. Table 3 lists, in order of Antimicrobial sensitivities of 106 obligately anaerobic isolates
prevalence, those bacteria isolated in large numbers from horses are presented in Table 5.
with pleuropneumonia.
Obligately anaerobic bacteria were recovered more commonly Discussion
from horses that died or were euthanased (16 of 19 horses; Our results are similar to those of studies in the USA, which
84%) than from those that survived (7 of 15 horses; 47%). In found pneumonia to occur predominantly in young, racing
11 of the horses, which died in the former group, therapy had animals, often Thoroughbreds.1,7 On the other hand, a study of
not been instituted until 2 d or more (range 2 to 21 d) after the 11 cases conducted in the UK reported show-jumpers to be
onset of clinical signs. Anaerobic bacteria were cultured from all most commonly affected. 5 These studies suggest that if
10 horses which had a putrid odour to their breath and/or Thoroughbreds are over represented, it is not due to breed per
pleural fluid sample (horses 67, 79, 107, 121, 124, 164, 175, se, but may be likely a reflection of their exposure to the stresses
242, 243, 261). Nine of these horses died or were euthanased which appear to be important in determining the susceptible

556 Aust Vet J Vol 78, No 8, August 2000

 Scientific

Table 5. Antimicrobial susceptibility of 106 obligately anaerobic bacteria from lower respiratory infection in 34 horses

Identity of isolate (No. tested) Proportion (%) of isolates susceptible to antimicrobial agent

Penicillin Metronidazole Amoxycillin Doxycycline Erythromycin

2U/mL 2 µg/mL 4 µg/mL 6 µg/mL 3 µg/mL

Peptostreptococcus anaerobius (7) 100 100 100 100 100

Other Gram positive cocci (22) 86 86 86 100 82
Clostridium spp (4) 75 100 75 100 100
Eubacterium fossor (11) 100 100 100 100 100
Other Gram positive nonsporing bacilli (16) 100 100 100 94 94
G- cocci (4) 75 100 100 100 25
Bacteroides fragilis (4) 0 100 0 100 100
Other bile+, indole- 100 83 100 100 100
Bacteroides tectum (6)
Indole+, nonpigmented 100 100 100 100 100
Prevotella heparinolytica spp (6)
Pigmented, asaccharolytic 78 100 78 100 100
Porphyromonas/Prevotella spp (9)
Nonpigmented, indole-, bile- 100 90 100 100 90
Prevotella spp (10)
Fusobacterium spp (7) 1 100 100 100 100

population. Recent transportation and long distance transporta- cultured from equine pleuropneumonia have been shown to be
tion were the most important risk factors identified by Austin et present as part of the oropharyngeal flora in normal horses,
al28 in their case-control study of development of pleuropneu- pulmonary samples must be collected without contamination
monia in horses and these findings reflected earlier studies.1,6 by oropharyngeal secretions.2,9 Thus, while endoscopic guarded
The most common stresses in our study were strenuous exercise tracheal swabbing has been used to obtain lower respiratory
and long-distance transportation (one or both preceding 76% samples TTA remains the preferred method by which to obtain
of cases). such samples from horses.32-34 In addition, the characteristics of
Delay in diagnosis (of as little as 24 h) and initiation of treat- pleural fluid samples obtained from left and right sides of the
ment, often led to an unfavourable outcome. Byars and Becht29 thorax were often different, indicating that bilateral thoracocen-
also found a more favourable outcome with early diagnosis and tesis should be performed. The inflammation and fibrin deposi-
early institution of treatment. In subsequent studies, Seltzer and tion associated with pleuritis may cause the normally fenes-
Byars30 found 61% of horses with uncomplicated pleuropneu- trated equine mediastinum to become complete, effectively
monia returned to racing. Less favourable findings have been dividing the pleural cavity.35
noted for complicated pleuropneumonia (for example In this study S equi subsp zooepidemicus, Pasteurellaceae (espe-
pulmonary abscesses) from which anaerobes were isolated.7 Our cially Pasteurella caballi and Actinobacillus suis-like isolates), E
study, in which 56% of horses died or were euthanased and coli, Bacteroides spp and Prevotella/Porphyromonas spp were
24% returned to racing, also showed a less favourable recovery isolated most frequently and in greater abundance. Such
rate for horses from which anaerobes were isolated. Many of the aerobic/facultatively anaerobic bacteria were isolated from both
individual cases were indeed complicated cases, drawing as they early and more advanced cases, but obligately anaerobic bacteria
did largely from university and specialist equine clinics. The were isolated in large numbers only after at least 5 d of illness.
distinctly poorer recovery rates with delay in treatment empha- In only two instances were E coli isolated within the first 48 h of
sizes the importance of taking all measures to obtain an early illness and then only via enrichment broth. In the other 12 cases
diagnosis. it was recovered after at least 7 d of illness – often accompanied
An early diagnosis of equine pleuropneumonia is aided by the by antimicrobial treatment, suggesting it may play a role in the
collection of appropriate samples from the lower respiratory chronic phase of pleuropneumonia. The progression of mixed
tract. This study illustrated the importance of obtaining a infections through several stages, characterised by a changing
sample for bacteriology that was representative of pulmonary bacterial population, has been documented in human clinical
lesions as well as pleural fluid. In several instances bacteria failed studies and experimental models of polymicrobial sepsis.36
to grow from pleural fluid samples whereas they could be Synergy between aerobic/facultatively anaerobic and anaer-
cultured readily from pulmonary samples (that is either TTAs obic bacteria has been documented in experimental infections.
from living horses or lung parenchymal samples taken at The mechanisms of such synergy are largely unknown but may
necropsy) from the same animal. These findings are consistent involve protection from phagocytosis and intracellular killing,
with those of Light et al31 who reported that pleural effusions production of essential growth factors and lowering of local
complicating pneumonia in humans may not yield bacterial oxygen concentration and oxidation and reduction potentials in
growth. In other cases, thoracocentesis yielded bacteriological host tissue. All these factors may thus provide appropriate envi-
results different (most often fewer isolates and/or no anaerobic ronmental conditions for replication and invasion by anaerobic
isolates) from those of pulmonary samples. Since bacteria components of infection.37

Aust Vet J Vol 78, No 8, August 2000 557

Scientific

The problem of distinguishing significant organisms (that is suggested that the underlying pathological processes progressed
those essential for disease or disease progression) among the very rapidly. These are obviously limiting factors in the treat-
potential pathogens, commensals and symbionts in a polymi- ment of equine pleuropneumonia. In one case (horse 164) in
crobial infection becomes a matter of practical concern when which pneumonia followed transportation together with racing,
considering antimicrobial treatment of such infections. Rather the lesions were particularly severe and more generalised than
than antimicrobial therapy directed at removing as many of the most of the other cases. This case was similar to those described
multiple species as possible, it may be as useful to treat strategi- as acute infarctive pneumonia.46 Such cases have been reported
cally selected components of the infection, thereby disturbing to be particularly refractory to treatment due to the over-
the microbial balance sufficiently to allow host defences to whelming acute lesions. Even where the lesions were more
control the disease process. Such an approach has been taken in localised, there were frequently large areas of cavitated necrosis.
the treatment of polymicrobial pulmonary infections in Importantly too, these were often adjacent to the mediastinum.
humans. The response of mixed pleuropulmonary infections in The internal location of such abscesses makes adequate
humans to antimicrobial therapy directed only against obli- drainage, an important therapeutic procedure in chronic cases,
gately anaerobic bacteria has been as good, if not better, than difficult and this contributes to the protracted course of these
that to broad spectrum antibiotic therapy for both aerobic and infections or, at worst, therapeutic failure.
anaerobic organisms.38 Many of these human infections were at Bacteria may gain entry to pulmonary parenchyma via the
a chronic stage of disease. In view of our findings that obligately blood from an extrapulmonary source, by inhalation of
anaerobic bacteria become a feature only after at least five days aerosolised bacteria and following aspiration of oropharyngeal
of illness, this approach could be unwise if the antimicrobial bacteria. The distribution of lesions in 14 of the 15 cases that
used in the acute phase did not have efficacy against the faculta- were necropsied suggested that infection gained entry to the
tively anaerobic species which predominate then. Penicillin has lungs under the effect of gravity and was consistent with aspira-
a good spectrum of activity against anaerobes and would also tion. In most cases pulmonary lesions were concentrated in
target the most prevalent facultative species S equi subsp zooepi- areas of lung supplied by those airways first to branch ventrally
demicus (recovered from 82% horses). However, its inability to in the standing horse.
penetrate effectively into purulent exudate is well docu- Many of the ‘stressful’ events associated with the onset of
mented.35,39,40 The presence of facultatively anaerobic Gram- pleuropneumonia in horses may not only interfere with
negative bacteria in 85% of cases would suggest the use of an pulmonary defenses, but also offer the potential for aspiration.
additional agent (Table 4). Many of the isolates in this study In humans, where aspiration is the most common route of
were susceptible to gentamicin in vitro although gentamicin is pulmonary infection, oropharyngeal secretions contain a very
also said to lack good penetration into respiratory tract secre- high concentration of bacteria (as many as 107 anaerobic and
tions. The broad-spectrum agent ceftiofur is considered to have 106 aerobic and/or facultatively anaerobic organisms per 0.1
good activity against many of the Gram-negative, facultatively mL) and therefore the aspiration of even small quantities of
anaerobic bacteria as well as S equi subsp zooepidemicus, which oropharyngeal secretions may present a very large bacterial
should make it a useful antibiotic for early cases.41 inoculum to the lung. In addition, the pulmonary defenses
Unfortunately, our studies predated the availability of this appear to be less efficient in handling aspirated bacteria than
antimicrobial. they are in eliminating inhaled, aerosolised organisms, although
Additional drugs such as metronidazole specifically directed the reasons for this are not understood.47
against anaerobes, may not be required if a diagnosis is made During transportation, horses are often cross-tied, unable to
and appropriate treatment instituted within 24 h of onset of lower their heads. This favours aspiration of pharyngeal material
illness. In addition, the prevalence of penicillin-resistant as well as placing the animal in a less favourable position for
bacteria among the 106 obligately anaerobic bacteria recovered drainage of lower respiratory secretions and coughing.12,47-49
during this study was very low (Table 5), with large numbers of During hard exercise there may be aspiration of pharyngeal
Bacteroides fragilis recovered from only one horse. secretions associated with extreme respiratory effort. 50
Metronidazole is more likely to be useful later in the disease. Racetrack debris has been observed in the tracheas of horses
In this study there was a great range of antimicrobial treat- after racing.51 In humans, a history suggestive of possible aspira-
ment regimens used among, and even within, individual cases at tion is an important clue to the possibility of anaerobic pleuro-
different stages of disease. This, together with the few samples pulmonary infection.52 All but five cases in the present study
obtained when prior antimicrobial treatment had not been had a history of possible aspiration. Thus, as with humans,
given, precluded meaningful examination of the influence of history is important in suggesting the possibility of lower respi-
antimicrobial therapy generally or of individual agents on the ratory infections in horses.
bacteriology of equine pleuropneumonia. This study certainly When oropharyngeal bacteria contaminate the lower respira-
showed that many of the organisms present in the diseased sites tory tract, their transfer to a new environment may aid potential
were not eliminated by the antimicrobial regimen instituted pathogens once they are removed from their normally
before investigations occurred. It was commonplace, particu- constrained niche among the normal flora of the pharynx.
larly in advanced cases, to culture bacteria that were sensitive, in Where the contamination is heavy, it is easy to see how
vitro at least, to antimicrobial therapy current at the time of pulmonary defenses may be overwhelmed. In some cases
sampling. This was not surprising, since both bacterial and however, where there may be stress-induced suppression of
tissue factors influence the results of antimicrobial pulmonary defenses, it may not necessarily require an abnor-
therapy.39,40,42-45 mally increased bacterial inoculum (via aspiration) to initiate
The lesions in cases that were necropsied were often infection. Those components of the normal pharyngeal flora,
extremely severe and the association between delayed onset of which are found as contaminants of the lower respiratory tract
treatment (as little as 24 h) and failure to recover from infection may be all that are required.18 These organisms rapidly become

558 Aust Vet J Vol 78, No 8, August 2000

 Scientific

involved in significant disease which may not be reversible if nolyticus sp nov isolated from humans with periodontitis. Int J Sys Bacteriol
1985;35:438-442.
treatment is not prompt and aggressive. 26. Love DN, Johnson JL, Jones RF, Bailey M. Comparison of Bacteroides
zoogleoformans strains isolated from soft tissue infections in cats with strains
Acknowledgments from periodontal disease in humans. Infect Immun 1985;47:166-168.
The work was supported in part by the NSW Racing 27. Love DN, Bailey GD, Collings S, Briscoe DA. Description of Porphyromonas
Research Fund. Media were prepared by Lana Patoka. circumdentaria sp. nov. and reassignment of Bacteroides salivosus (Love,
Johnson, Jones, and Calverley 1987) as Porphyromonas (Shah and Collins
1988) salivosa comb. nov. Int J Syst Bacteriol 1992;42:434-438.
References 28. Austin SM, Forman JH, Hungerford LL. Case control study of risk factors for
1. Raphel CF, Beech J. Pleuritis secondary to pneumonia or lung abscessation development of pleuropneumonia in horses. J Am Vet Med Assoc
in 90 horses. J Am Vet Med Assoc 1982;181:808-810. 1995;207:325-328.
2. Bailey GD, Love DN. Oral associated bacterial infections in horses: studies 29. Byars D, Becht JL. Pleuropneumonia. Vet Clin North Am Equine Pract
on the normal anaerobic flora from the pharyngeal tonsillar surface and its asso- 1991;7:63-67.
ciation with lower respiratory tract and paraoral infections. Vet Microbiol 30. Seltzer KL, Byars TD. Prognosis for return to racing after recovery from
1991;26:367-379. infectious pleuropneumonia in Thoroughbred racehorses:70 cases (1984-
3. Smith BP. Pleuritis and pleural effusion in the horse: a study of 37 cases. J 1989). J Am Vet Med Assoc 1996;208:1300-1301.
Am Vet Med Assoc 1977;170:208-211. 31. Light RW, Girard WM, Jenkinson SG, George RB. Parapneumonic effu-
4. Biberstein EL, Knight HD, England K. Bacteroides melaninogenicus in sions. Am J Med 1980;69:507-512.
diseases of domestic animals. J Am Vet Med Assoc 1968;153:1045-1049. 32. Sweeney CR, Sweeney RW, Benson CE. Comparison of bacteria isolated
5. Mair TS, Lane JG. Pneumonia lung abscesses and pleuritis in adult horses: a from specimens obtained by use of endoscopic guarded tracheal swabbing and
review of 51 cases. Equine Vet J 1989; 21: 175-180 . percutaneous tracheal aspiration in horses. J Am Vet Med Assoc
6. Bernard-Strother S, Mansmann RA. Diagnosis and treatment of anaerobic 1989;195:1225-1229.
bacterial pleuropneumonia in six horses. Compend Contin Educ Pract Vet 33. Darien BJ, Brown CM, Walker RD, Williams MA, Derksen FJ. A tracheo-
1985;7:S341-S349. scopic technique for obtaining uncontaminated lower airway secretions for
7. Sweeney CR, Divers TJ, Benson CE. Anaerobic bacteria in 21 horses with bacterial culture in the horse. Equine Vet J 1990;22:170-173.
pleuropneumonia J Am Vet Med Assoc 1985;187:721-724. 34. Christley RM, Hodgson DR, Rose RJ, Reid SWJ, Hodgson JL. Comparison
8. Sweeney CR, Holcombe SJ, Barningham SC, Beech J. Aerobic and anaer- of bacteriology and cytology of tracheal fluid samples collected by percutaneous
obic bacterial isolates from horses with pneumonia or pleuropneumonia and transtracheal aspiration or via an endoscope using a plugged, guarded
antimicrobial susceptibility patterns of the aerobes. J Am Vet Med Assoc catheter. Equine Vet J 1999;31:197-202.
1991;198:839-842. 35. Schott HC, Mansmann RA. Thoracic drainage in the horse. Compend
9. Bailey GD. Studies on oral-associated bacterial infections in horses [PhD Contin Educ Pract Vet 1990;12:251-261.
thesis]. University of Sydney, 1989. 36. Gorbach SL. The preeminent role of anaerobes in mixed infections. J
10. Collins M B, Hodgson DR, Hutchins DR. Pleural effusion associated with Antimicrob Chemother 1982;10 (Suppl A):1S-6S.
acute and chronic pleuropneumonia and pleuritis secondary to thoracic wounds 37. Brook I. Encapsulated anaerobic bacteria in synergistic infections. Microbiol
in horses: 43 cases (1982-1992) J Am Vet Med Assoc 1994;205:1753-1758. Rev 1986;50:452-457.
11. Reimer JM, Reef VB, Spencer PA. Ultrasonography as a diagnostic aid in 38. Finegold SM, Bartlett JG, Chow AW et al. Management of anaerobic infec-
horses with anaerobic bacterial pleuropneumonia and/or pulmonary abscessa- tions. Ann Intern Med 1975;83:375-389.
tion:27 cases (1984-1986). J Am Vet Med Assoc 1989;194:278-282. 39. Pennington JE. Penetration of antibiotics into respiratory secretions. Rev
12. Anon. Equine roundtable discussion: respiration (part 3). Equine Pract Infect Dis 1981;3:67-74.
1989;11:29-39. 40. Dixon PM, Thomson JR, Brown R, Galbraith A, Bogan JA. Some observa-
13. Edwards DJ. Transtracheal aspiration - a technique using reusable mate- tions on the penetration of antimicrobial drugs into the respiratory secretions of
rials. Aust Equine Vet Assoc Newsletter 1988;6:34-35. horses. In: Lung Function and Respiratory Diseases in the Horse. Deegen E,
14. Bailey GD, Love DN. Eubacterium fossor sp. nov., an agar-corroding Beadle RE. eds. International Symposium, Hannover Germany 1985;35-37.
organism from normal pharynx and oral and respiratory tract lesions of horses. 41. Prescott JF. Antimicrobial drug selection for lower airway infection In:
Int J Syst Bacteriol 1986;36:383-387. Robinson N E (editor) Current Therapy in Equine Medicine 4. W B Saunders,
15. Bailey GD, Moore LVH, Love DN, Johnson JL. Bacteroides heparinolyticus: Philadelphia, 1997:454-459.
deoxyribonucleic acid relatedness of strains from the oral cavity and oral-asso- 42. Beech J. Therapeutic strategies involving antimicrobial treatment of the
ciated disease conditions of horses, cats and humans. Int J Syst Bacteriol lower respiratory tract in large animals. J Am Vet Med Assoc 1984;185:1199-
1988;38:42-44. 1202.
16. Love DN, Johnson JL, Jones RF, Bailey M, Calverley A. Bacteroides tectum 43. Bazra M. Principles of tissue penetration of antibiotics. J Antimicrob
sp nov and characteristics of other nonpigmented Bacteroides isolated from Chemother 1981;8 (Suppl):7S-28S.
soft-tissue infections from cats and dogs. Int J Syst Bacteriol 1986;36:123-128. 44. Schlessinger D. Failure of aminoglycoside antibiotics to kill anaerobic, low
17. Racklyeft DJ. Studies on bacterial infections of the equine lower respiratory pH and resistant cultures. Clin Microbiol Rev 1988;1:54-59.
tract [PhD thesis] University of Sydney, 1990. 45. Sahn SA, Taryle DA, Good JT Jnr. Experimental empyema: time course and
18. Racklyeft JD, Love DN. Bacteriology of normal lower respiratory tract of pathogenesis of pleural fluid acidosis and low pelural fluid glucose. Am Rev
horses and its significance for lower respiratory tract disease. Aust Equine Vet Resp Dis 1979;120:355-361.
1999;17:66-75. 46. Wilson WD. Pleuropneumonia and acute infarctive pneumonia In:
19. Bauer AW, Kirby WMM, Sherris JC, Turck M. Antibiotic susceptibility testing Proceedings of 17th Bain-Fallon Memorial Lectures. Australian Equine
by a standardized single disk method. Am J Clin Path 1966;45:493-496. Veterinary Association 1995:297.
20. Wilkins TD, Thiel T. Modified broth-disk method for testing antibiotic suscep- 47. Finegold SM. Aspiration pneumonia, lung abscess and empyema. In:
tibility of anaerobic bacteria. Antimicrob Agents Chemother 1973;3:350-356. Pennington JE, editor. Respiratory Infections: Diagnosis and management .
21. Schlater LK, Brenner DJ, Steigerwalt AG et al. Pasteurella caballi, a new Raven Press, NY, 1983:191-199.
species from equine clinical specimens. J Clin Microbiol 1989;27:2169-2174. 48. Racklyeft DJ, Love DN. Influence of head posture on the respiratory tract of
22. Bisgaard M, Piechulla K, Ying Y-T, Frederiksen W, Mannheim W. healthy horses. Aust Vet J 1990;67:402-405.
Prevalence of organisms described as Actinobacillus suis or haemolytic 49. Raidal SL, Love DN, Bailey GD. Inflammation and increased numbers of
Actinobacillus equuli in the oral cavity of horses. Comparative investigations of bacteria in the lower respiratory tract of horses within 6 to 12 hours of confine-
strains obtained and porcine strains of A suis sensu stricto. Acta Path Microbiol ment with the head elevated. Aust Vet J 1995;72:45-50.
Scand Sect B Microbiol Immunol 1984;92:291-298. 50. Raidal SL, Love DN, Bailey GD. Effect of a single bout of high intensity exer-
23. Jang SS, Biberstein EL, Hirsh DC. Actinobacillus suis-like organisms in cise on lower respiratory tract contamination in the horse. Aust Vet J
horses. Am J Vet Res 1987;48:1036-1038. 1997;75:293-295.
24. Holdeman-Moore LV, Johnson JL, Moore WEC. Genus Peptostreptococcus. In: 51. Arthur RM. Subacute and acute pleuritis. Proc Am Assoc Equine Pract
Sneath PHA, Mair NS, Sharpe ME, Holt JG, editors Bergey’s Manual of 1983;29:65-69
Systematic Bacteriology, Volume 2. Williams and Wilkins, Baltimore, 52. Bartlett J G, Finegold SM. Anaerobic infections of the lung and pleural
1986:1083-1092. space. Am Rev Respir Dis 1974;110:56-77.
25. Okuda K, Kato T, Shiozu J, Takazoe I, Nakamura T. Bacteroides hepari- (Accepted for publication 3 April 2000)

Aust Vet J Vol 78, No 8, August 2000 559