Pathologic patternsof Serratia marcescens pneumonia

PATHOLOGIC PATTERNS 0 F SERRATIA MARCESCENS PNEUMONIA Jeffrey D. Goldstein, MD,*JohnJ. Godleski, MD,? Jerry P. Balikian, MD,$ and Peter G. Herman, MDw

To characterize the pulmonary lesions caused by S e r r a t i a m a r c e s the authors reviewed all autopsy-culture-proven cases of S. m a r c e s c e n s pneumonia occurring at their hospital between 1968 and mid-1980. In 16, S . m a r c e s c e n s was the only organism cultured from the lungs during life or at autopsy. This report describes primarily these pure infections. Two histopathologic reactions were seen. Nine non-neutropenie patients had acute, hemorrhagic bronchopneumonia, seven with microabscesses and two with larger cavities. In seven, distinctive vasculitis was apparent in vessels larger than 75 microns in diameter; intramural gram-negative rods were identified in two. Seven immunosuppressed patients had diffuse neutropenic pneumonitis resembling diffuse alveolar damage, with extensive intra-alveolar fibrinous exudates and pulmonary hemorrhage. In two patients, bacteria without cellular reaction were present. In patients with prolonged infections, focal areas of intra-alveolar organization and bronchiolitis obliterans accompanied both patterns. Since the incidence of nosocomial S. m a r c e s c e n s infection is increasing and since pneumonia caused by this organism is recognizable histologically, autopsy cultures positive for S . m a r c e s c e n s should not be disregarded. Hum Pathol 13:479-484, 1982.

METHODS

cens,

a g r a m - n e g a t i v e r o d bacterium once t h o u g h t harmless, has been recognized as an i m p o r t a n t cause o f nosocomial infections. Several hospitals, including o u r own, have e x p e r i e n c e d epidemics orS. marcescens infection; frequently, these have been associated with respiratory therapy e q u i p m e n t , t-7 Pathologic features o f S. m a r c e s c e n s p n e u m o n i a previously have been discussed only in case reports. 8 An o n g o i n g epidemic o f S. marcescens infection p r o m p t e d us to characterize the pathologic patterns typical o f this bacterium. T h e clinical and radiographic features o f some o f the affected patients trove been r e p o r t e d previously. ~ Serratia marcescens,

Accepted for publication August 17, 1982. Dr. Goldstein is supported in part by National Research Service Award 4-T32ttLO7066-05. This paper was presented at the Seventieth Annual Meeting of the US-Canadian Division of the International Academy of Pathology, Chicago, Ill, March 1981. * Research Fellow, Departments of l'athology, Briglmm and Women's ttospital and Harvard Medical School, Boston, Mass. [" Pathologist, Brigham and Women's Hospital, and Assistant Professor of Pathology, Harvard Medical School, Boston, Mass. :~Director, Division of Oncoradiology, Sidney Farber Cancer Institnte, and Associate Professor of Radiolog% tlarvard Medical School, Boston, Mass. wDirector, Division of Thoracic Radiology, Brigham and Women's ttospital, and Professor of Radiology, Harvard Medical School, Boston, Mass. Address correspondence and reprint requests to Dr. Goldstein, Department of Pathology, Brigham and Women's Hospital, 75.Francis St, Boston, MA 02115.

C u l t u r e - p r o v e n cases o f S . marcescens p n e u m o n i a occurring at o u r hospital between 1968 and August 1980 w e r e r e t r i e v e d by c o m p u t e r i z e d s e a r c h o f SNOP(Systetnized Nomenclature of Pathology)coded autopsy diagnoses and by manual review o f autopsy protocols for cases a f t e r 1978, which were not yet coded. All autopsy lung slides, gross descriptions and photographs, clinical charts, and bacteriologic records were studied. Data collected by the hospital infection-control unit were reviewed to "determine how many patients had had S. marcescens infections since N o v e m b e r 1977.

RESULTS

Forty cases were f o u n d , 30 o c c u r r i n g ariel" 1977, c o r r e s p o n d i n g to a clinical o u t b r e a k o f S. marcescens (fig. 1). T h e s e cases were divided into t h r e e groups. G r o u p I (16 patients) consisted o f cases in wlfich S. marcescens was the sole organism cultured f r o m respiratory secretions b e f o r e death and from tile lung at atttopsy, o r at autopsy alone, if p r e - m o r t e m s p u t n m cultures w e r e negative o r were not p e r f o r m e d . In g r o u p 2 (six patients), organisms in addition to S. marcescens were cultured f r o m s p u t u m b e f o r e death, but only S e r r a t i a organisms were identified at the time o f autopsy. In g r o u p 3 (18 patients), m o r e than one bacterial o r fungal organism was identified in tile lungs postmortem. T h r e e patients in tiffs study had cells containing Cowdry type A i n t r a n u c l e a r inclusions suggestive o f H e r p e s s i m p l e x infection. T h e s e possible herpetic lesions were not sufficiently widespread to alter the histologic a p p e a r a n c e o f the bacterial infection. In one case, intranuclear inclusions were limited to the trachea; in the o t h e r two, p u h n o n a r y necrosis was not present. P o s t m o r t e m viral cultures were obtained, and electron microscopy was p e r f o r m e d in o n e o f the cases in which inclusions were seen in the hmg, but these failed to yield evidence o f viruses. ThereforCr, 'two o f these t h r e e patients were included in g r o u p 1 and the o t h e r was included in g r o u p 2, on the basis of" the criteria listed above. Unless otherwise stated, the pathologic features described were those o f the 16 p u r e infections o f g r o u p 1.

0046-8177/82/0500/0479 S I.20 ~ W. B. Saunders Co.

479

HUMAN I'ATHOLOGY--VOLUME 13, NUMBER 5

May 1982

in a n u r s i n g h o m e , a n d o n e in a n o u t p a t i e n t h e m o d i a l y s i s unit. T h e final hospitalization lasted between o n e a n d 90 days (median, 19.5 days). T h e patients had generally r e c o g n i z e d risk factors for nosocomial infection: five had e n d o t r a c h e a l tubes, seven had u r i n a r y catheters, a n d antibiotics were a d m i n i s t e r e d p r i o r to the onset o f the S. m a r c e s c e n s i n f e c t i o n in twelve. T h e r e s p i r a t o r y tract o r b l o o d s t r e a m a p p e a r e d to be the m o s t c o m m o n e n t r y portal: only two patie.nts had w o u n d infections Rrior to acquiring p n e u m o n i a , and n o n e h a d initial bactnria. On the o t h e r b a n d , the first positive clinical c u h u r e f r o m s e v e n p a t i e n t s was e i t h e r b l o o d o r sputttm. T h r e e patients had b a c t e r e m i a s i m u l t a n e o u s witlt o r s u b s e q u e n t to respiratory infection. S . m a r c e s c e n s grew in p r e m o r t E m blood c u h u r e s but not in s p u t u m fi'om t h r e e additional patients; tiffs m a y re[1ECt both the difficulty in obtaining a d e q u a t e specim e n s f r o m the r e s p i r a t o r y tract a n d tire rapid dissemination o f tiffs o r g a n i s m into the b l o o d s t r e a m . In seven patients S . m a r c e s c e n s was not c u l t u r e d b e f o r e death; t h r e e o f these had both positive l u n g a n d positive blood cultures at autopsy. T h e ages a n d u n d e r l y i n g diseases o f the 24 patients with m i x e d infections (groups 2 a n d 3) were similar to those o f patients with the p u r e infections (table 1). A l t l m u g h the length o f hospitalization was longer for patients in g r o u p s 2 a n d 3 - - m e a n , 35.9 days -4- 29.5 (SD); m e d i a n , 29 days vs. 24.8 days 26.5 (SD) a n d 19.5 days--th[~ d i f f e r e n c e is not statistically significant (unpaircd-i test, p > 0.10). A variety o f o t h e r , f r e q u e n t l y multiple, o r g a n i s m s were identified in t h e s e p a t i e n t s , i n c i n d i n g S t a p h y l o c o c c i , Klebsiella, P s e u d o m o n a s , E n t e r o b a c t e r , P r o t e u s , Candida, a n d Aspergillus species.

7

09

hi 03 t) It. O tW W ira Z

1968 '69"/'0'71 '72'73'74'75176L77 ~ ~919{)0 YEAR Figure 1. Distribution of fatal Serratia marcescens pneumonia

by )'ear. Group 1 = pure S. marcescens by pre- and post-morte,n culture. Group 2 = mixed infection premortem but otfly S. marcescens post-mortem. Group 3 = mixed infection post-mortem.

CLINICAL

FEATURES

Clinical f e a t u r e s a r e s u m m a r i z e d in table 1. Seven m e n a n d nine w o m e n had p u r e infections, ntaking u p g r o u p 1. Virtually all were elderly, imntunocom, promised, o r severely debilitated. Twelve patients had ntalignancies, nine of which were ttematologic. T w o patients were postoperative; one patient had c o r o n a r y a r t e r y disease, o n e lind ingEstEd a d r u g overdose, a n d one had cin'onic renal faihtre. T h e m e a n age o f the patients was 56.6 years --- 24.8 (SD). ElevEn patients w e r e m o r e than 50 )'ears o f age; t i r e five y o u n g e r patients had hematologic disorders (four) o r brain injury li'om d r u g o v e r d o s e (one). AItltough all cases were nosocomial, tltree arose outside o u r hospital: o n e originated in a n o t h e r hospital, one T A B L E 1.

PATHOLOGIC

FEATURES

OF PURE

SERRATIA MARCESCENS I N F E C T I O N S

1

T w o pathologic p a t t e r n s were seen, d e p e n d i n g on the host's d e f e n s e mechanisms. T h e m o r e corn-

C L I N I C A L F E A T U R E S O F P A T I E N T S W I T H SERRATIA MARCESCENS P N E U M O N I A

Total patients Men Women Mean age (years • SD) Mean duration of final hospitalization (days • SD)

Group t

Group 2

Group 3

16 7 9 56.6 • 23.4

6

18 11

59.6 -- 27

55.9 +- 16.8

24.8 • 26.5

27.3 - 18.7

38.8 -'- 32.3

9 3 2 1

l 0 2 1

6 2 6 0

2

3

1

1 1 0

0 0 O

2 2 2

3

1

3

6 0

7

Underlying Illness:

Hematologic malignancies and related disorders Nonheinatologic malignancy Postoperative Coronary artery disease Chronic ueurologic disease or acute brain damage Chronic renal failure or renal transplant Liver disease Burn or toxic epidermal necrolysis Patients with more than one predisposing condition

480

S E R R A T I A M A R C E S C E N S PNEUMONIA--GoH)srrnN ~:T AL.

Figure 2. Left, nccrotizing bronchopneumonia with abscess formation. The alveolar cxudate consists of neutrophils, macrophages, fibrin, and focal erythrocytes. (Original magnification • 16.) Right, gross photograph of a lung from a non-ncutropenic patient with S. marcescens pneumonia, demonstrating patchy consolklation from acute necrotizing bronchopneumonia. Ahhough upper and middle lobe infection is more extensive in this case, most patients had prominent involvement of the hmg bases. mon reaction, f o u n d in nine n o n - n e u t r o p e n i c patients, was an acute, necrotizing b r o n c h o p n e u m o n i a (fig. 2, left). Grossly, the lungs o f these patients showed patchy areas o f consolidation and focal hemorrhage, usually bilateral and most extensive in the lower lobes (fig. 2, right). Occasionally, b r o n c h o p n e u m o n i a became confluent. Seven patients had microabscesses, and two had larger cavities. These larger cavities appeared to be infected p u l m o n a r y infarcts in a patient with p u h n o n a r y embolism. L u n g weights ranged from 340 to 1650 g, with a mean o f 810 g - 4 0 5 (SD). T h r e e patients had pleural effusions in a m o u n t s greater than 150 ml. Microscopically, the alveolar exudates were observed to contain polymorphonnclear leukocytes and macrophages, fibrin, and various degrees o f hemorrhage. Bacilli were demonstrable morphologically in eight cases. Microabscess formation was d e f i n e d by alveolar septal necrosis and lysis o f the cellular infiltrate. T h e inIlammatory exudate extended into airways in five cases. Seven cases showed a distinctive vasculitis affecting both arteries and veins larger than 75 microns in dianteter (fig. 3). Neutrophilic infiltrates were localized to the intima or media; gramnegative rods were identified intramurally in two. Actual vascular necrosis was not seen, and associated thrombosis was found in only one case--that o f the patient with p u h n o n a r y emboli. Cases lacking vasculitis had earlier, less extensive b r o n c h o p n e u m o n i a . Seven neutropenic patients, including two bone m a r r o w transplant recipients, all with u n d e r l y i n g hematologic disorders, had a diffuse n e u t r o p e n i c pneumonitis. T h e hmgs were unifornfly edematous a n d h e m o r r h a g i c macroscopically. L u n g weights rfinged from 470 to 1500 g, with a mean o f 880 g +

341 (SD). None had significant pleural effusions. Microscopically, extensive f i b r i n o u s e x u d a t e s and h e m o r r h a g e filled the alveolar spaces (fig. 4, left). In two cases, bacteria without cellular.reaction were present (fig. 4, right). Alveolar macrophages were increased in several cases; some were hemosiderin-laden. Four patients had focal hyaline membranes. Generally, these cases resembled diffuse alveolar d a m a g e with other etiologies. Interstitial leukemic infiltrates were evident in four cases. In four patients, focal areas o f intra-alveolar organization and bronchiolitis obliterans developed; this a c c o m p a n i e d both necrotizing b r o n c h o p n e u monia and n e u t r o p e n i c pneumonitis (fig. 5). T h e most extensive organization occurred in two patients from group 3 whose hospitalizations lasted three and four months respectively. T h e i r hmgs showed extensive architectural remodeling and microscopic honeycombing. T h e six patients in g r o u p 2, none o f whom were leukopenic, all had acute b r o n c h o p n e u m o n i a similar in appearance to that seen in the non-neutropenic patients in group 1. Four had microabscesses and three had the vasculitis. DISCUSSION

Tile sanguine history o f S. marcescens has been reviewed previously. ~'1~ T h i s organism is now responsible for serious hospital-acquired infections a l d is frequently resistant to muhiple antimicrobials. I-7 Point-source outbreaks o f S. marcescens have been traced to a variety o f contaminated solutions including disinfectants, intravenous solutions, water used in respiratory care devices, and hand lotions? Medical e q u i p m e n t a n d m a n y diagnostic a n d t h e r a p e u t i c

481

HUMAN PATHOLOGYIVOLUME 13, NUMBER 5 May 1982

ii ~'~-

- --., ~ql.~ "

" -=/,~i'=i.~ll~42~ll i r ~ w l , - " r ~ ' , i .

Ii

~

~ .

r

"~

i r , ~ . . , e~i

r.lli l

~ ._...

. . . .

"i-'

,

,

. .~

.

i

.

9- i .

i~

.'.:).~

~.

'.,it,

'"

~.

" :

-

~:...

~

~*

-

";.'l~.' ,l~ -"

.:,~.,,,O~

.

~ql~. ~

.

.

.

~, z , ~ . . = _ - - . - -

.

.

9

-:-"

_lie

....

, , ..

=.

+

m ,.-

...._

,ll-"~"~-~, -: " - ' .

=.., ~12

~, . .

IL.--..~e .e;,l/~liT- - I L L _ ~ v

..

r

~-

-"~.... _:..:.

. .~.

.~

~

..

..,~l~:,....

~.-.,

]'iq~-.i,--.,~'l~

i:l

, - "" ,

"~

~ e . . . .

-i ~ - ' ; ~ , O - - - " ~ l

....

,l"..-'~ii-.~-

tl ~

v_

"o

l e "

"-

e` "~

*: ~,.'.',~ll," . . . , i ~ - ~ . l l t

"

"

. . l ~ ,

i

.: ~

l ! ' 7 " T e ~ ' . -"

; '~.~ ; l ~ l l ~ l l l r ' i i ' - ' 7 ' q l "

~-. , ' ~ i i

.,*

~'

~.,,.

,.

"

~

,~,~.,._ ...

,,,

~:.'" -

"ll

il]i

,i " ~# ~ .ai~ _r._i!3t ~'

IP

9

~ :'il~.,'~-

,,

,"Ji'~'~ - _ :'.~i i l

. ,i, . , J i , ~

.,i,.

. '. '-

~i~.., ,~,

"AT.---. "~ii'.,~-82

~

~ ~

i

. . . . " "~ " - " ~ -~* - . ;, -~' ". i ' ~ , ' - ; " :,:'~.-~-~.-.,~ o ~ - 4 1 ~" ~l i ~ , ~ . . ,' ~ " ~

'

--.--~',,~ ".

_.... ,

:*..}~ i ~ ' - i ~ , :

.

,..o

.

"~

; . p -~,

/ il;-,-~,.

. .~.,.,,,.~..J. ~ , i

.~J

.J ~ , . q l ' i l ' ~ i l t l t Ill

,,

-~-.

,!

. . . i ~ . . . . . .~.'~,-

-...~

F < "'i , ~ ' ' i r ,.~ir .... .-~-~*'-'A , #".

".~

~

_'l: ,'~P~' I

-;.~ill, ,"~. "." - ' - L ~ j ~ L ~ , , / .

".~I_~-_

, ? j...

,,~*..

~_~l'r.-

"-" .,,.

,~qlli_~lird~'ii~'~-~..

,~

F i g u r e 3 (top). Vasculitis i n v o l v i n g a n after)'. A c u t e i n i l a m m a t o r ) " ceils a r c p r e s e n t fi)cally in t h e s u b c n d o t h c l i a l r e g i o n a n d m e d i a . M u r a l necrosis and thrombosis are not evident. (Original niagnitication • F i g u r e 4. Bottom hft, d i f f u s e a l v e o l a r d a m a g c a n d h c v n o r r l i a g c f i ) u n d in n c u t r o p c v i i c p a t i e n t s w i t h S. marcescens i m e u m o n i a . Focal l i y a l i n e m e m b r a n e s ;ire p r e s e n t . ( O r i g i n a l m a g n i f i c a t i , i n • 16.)Bottom right, c l u s t e r s o f bacilli w i t h o u t n c u t r o p h i l i c r e a c t i o n . P u l m o n a r ) ' h e m o r r h a g e , h c m o s i d e r i n - l a d c n m a c r o p h a g c s , a n d b y p e r l l l a s l i c t y p e II p h c u m o c y t c s :ire a l s o p r e s e n t . ( O r i g i n a l m a g n i t i c a t i o n x 160.)

' p r o c e d u r e s m a y also d i s s e m i n a t e tile o r g a n i s m . Point-source outbreaks, however, are responsible for a relativel), small n u m b e r of cases. Cross-contamination o f patients by hospital personnel is the most important route o f transmission. T h e incidence 6f S. marcescens infections varies widely. Figures as high as 17 p e r cent have b e e n reportecl in n o s o c o m i a l pneumonias, ~ but in one stucly o f 52 episocles o f pneumonia in ]eukenlic patients, no cases were rec-

482

ognized despite all incidence o f other gram-negative bacilli exceeding 50 per cent. ~z Tile most serious l)reviously reportcd nosoconliai outbreak ofS. marcescens occurred in 1973 anti 1974 ill Nashville, Txznn: 210 patients were affcctccl in four hospitals? Between November 1977 and J u n e 1980, 600 patients at o u r hospital had positive bacteriologic cuhures f o r S. marcescens; in 242 o f these the organism was isolated fi'om respirator)' tract specimens. AI-

S E R R A T I A M A R C E S C E N S PNEUMONIA--GoLDSTEt.~ ET AL,

Figure 5. Bronchiolitis obliterans and intra-alveolar fibrosis. Areas of organization accompanied both necrotizing br0nchopneumonia and neutropenic pneumonitis in patients with prolonged S. marcescens infection. Oxygen toxicity may lmve contributed to the fibrosis, which was most extensive in two patients with mixed infections whose hospitalizations lasted 3 to 4 months. t h o u g h from the records available we were unable to distinguish between colonization a n d infection in these patients, the morbidity o f this o u t b r e a k has b e e n g r e a t . T h i r t y p a t i e n t s were f o u n d to h a v e c u h u r e - p r o v e n S. marcescens p n e u m o n i a at autopsy. T h e total mortality from S. marcescens a m o n g these 600 patients is not known. T h e m a g n i t u d e o f tiffs o u t b r e a k and the relatively large n u m b e r o f cases e x a m i n e d p o s t m o r t e m permits a definitive description o f the pathologic f e a t u r e s o f S. m a r c e s c e n s p n e u m o n i a . We h a v e g r o n p e d these cases so that o u r descriptions are based primarily on p u r e infections as d e t e r m i n e d by both p r e m o r t e m and p o s t m o r t e m bacteriologic cultures ( g r o u p 1). Tiffs was necessary because m o r e than half the patients had multiple organisms, which might alter the lfistologic reaction. In actuality, some o f the cases we included in g r o u p s 2 and 3 may have been p u r e S. marcescens infections as well, since the o t h e r organisms f o u n d may have been airway colonizers r a t h e r t h a n p u l m o n a r y p a t h o g e n s . H o w e v e r , we excluded these cases fl'om o u r description to eliminate any possibility that the p n e u m o n i c process resulted f r o m an organism o t h e r than S. marcescens. We recognized two d i f f e r e n t patterns o f ptdmon a r y i n j u r y in S. marcescens infection. An a c u t e , h e m o r r i m g i c , necrotizing b r o n c h o p n e u n m n i a with microabscesses was most c o m m o n . A distinctive featttre o f this pattern was a vasctditis in at least scattered arteries and veins in 75 p e r cent o f cases. This angiitis d i f f e r e d f r o m that caused by Pseudomonas aeruginosa ~3 in that i n t r a m u r a l organisms were rare, acute int l a m m a t o r y cells were focal o r eccentrically distributed in the stlbendothelial region o f the vessel wall,

and thrombosis o r vascular necrosis were unnsual. A similar "septic" vascnlitis recentl}~' has been described in 30 per cent o f cases o f Legionella pneumonia./4 Vasculitis is not a recognized c 6 m p o n e n t o f Klebsiella, Proteus, o r Escherichia coli p n e u m o n i a . 15-19 This vasculitis suggests that helnatogenous s p r e a d o f the organism may occur shortly after infection begins, thus explaining the high f r e q u e n c y o f bacteremia seen simultaneously with o r shortly after the onset o f pneumonia. Immunosuppressed patients incapable of m o u n t i n g a neutrophilic response had an agranulocytic o r n e u t r o p e n t c pneumonitis that grossly and microscopically reflected diffuse alveolar damage. 2~ P u h n o n a r y h e m o r r h a g e was p r o m i n e n t , b u t organisms were identified in only two cases. Cellular responses to infections are generally missing in patients with helnatologic malignancy, 2~ and fungal and Pneumocystic carinii infections in i m m u n o s u p p r e s s e d patients previously had shown only diffuse alveolar d a m a g e / z Although both g r o u p 1 patients who had intranuclear inclusion bodies exhibited this pattern, the viral infection did not a p p e a r substantial e n o u g h to explain the histologic appearance. T h e coexistent finding o f p u l m o n a r y leukemic infiltrates in four o f the i m m u n o s u p p r e s s c d patients is not surprising; alone, the presence o f these infihrates is insufficient to cause respiratory symptoms. 23 Even t h o u g h o u r pathologic descriptions did not include the mixed infections, we r e p o r t these cases for several reasons. First, they help d o c u m e n t the m a g n i t u d e o f the S. marcescens o u t b r e a k and d e m o n strate that patients at risk for nosocomial infections f r c q n e n t l y may be affected by m o r e than one 0r-

483

HUMAN PATHOLOGY--VOLUME

13, N U M B E R 5

May 1982

ganism. Second, they emphasize the difficulty in eradicating pathogens resistant to multiple antibiotics, since group 2 patients succumbed to S. m a r c e s c e n s after other organisms were apparently controlled. Finally, because patients who had S . m a r c e s c e n s pneumonia the longest inevitably became infected by other organisms, the chronic intra-alveolar organization and extensive puhnonary fibrosis seen in some of their hmgs could not be attributed solely to S. m a r c e s c e n s infection. These dramatic changes, however, most likely resuhed from the continued presence of S. marcescens in the lung b e c a u s e o f ineffective antimicrobial therapy. Chronic oxygen toxicity may have had a secondary role in these cases. Prevention o f nosocomial S. m a r c e s c e n s infection requires all_ infection surveillance program. The autopsy can be a vital procedure for identifying S. m a r cescens infection and instituting crucial prophylactic measures. In eight group 1 patients (50 per cent), S. m a r c e s c e n s was not suspected clinicall)'. In two of these patients, premortem cuhures eventually grew S, marcescens after death; in the remaining six, the organism was first found at autopsy. An awareness of the histologic manifestations of this organism in the lung should allow pathologists to recognize S. m a r c e s c e n s pneumonia. Autopsy cnltures positive for this organism should not be attributed to mere colonization or postmortem contamination. ACKNOWLEDGMENTS

The attthors thank Mary Sweeney, RN, and Ira Tager, MD, of the Brigham and Women's Hospital Infection Control Unit for making their data available tO US.

REFERENCES i. Yu VL: Serratia marcescens: historical perspective and clinical review. N Engl J Med 300:887, 1979. 2. Sanders F Jr, Luby JP, Johanson WH Jr, et al.: Serratia mattestens infections from inhalation therapy medications: nosocomial outbreak. Ann Interu Med 73:15, 1970.

484

3. Cabrera HA: An outbreak ofSerratia marcescens and its control. Arch Intern Med 123:650, 1969. j 4. Ringrose RE, McKown B, Felton FG, et al: A hospital outbreak of Serratia marcescens associated with ultrasonic nebulizers. Ann Interu Med 69:719, 1968. 5. Schaberg DR, Afford PH, Andersma R, et al: An outbreak of nosocomial infection due to multiply resistant Serratia marcescens: evidence of interhospital spread. J Infect Dis 134:181, 1976. 6. Wilkowske CJ, Washington JA, Martin WJ, et al: Serratia marcescens: hiochemical characteristics, antibiotic susceptibility patterns, and clinical significance. J AMA 214:2157, 1970. 7. David JT, Foltz E, Blakemore WS: Serratia marcescem: a pathogen of increasing clinical importance. JAMA 214:2190, 1970. 8. Mehz DJ, Grieco MH: Characlerlstics of Serratia marcescens pneumonia. Al~h Intern Med 132:359, 1973. 9. BalikianJP, Herman I'G, and GodleskiJJ: Serratia pneumonia. Radiology 137:309, 1980. 10. Gaughran ERL: From superstition to science: tile history of a bacterium. Trans NY Acad Sci 31:3, 1969. I1. Lorber B, Swenson RM: Bacteriology o f aspiration pqeumoqia: a prospective study of community and hospital acquired cases. Ann Intern Med 81:329, 1974. 12. Sickles EA, Young VM, Greene WH, et al: t'neumonia in acute leukemia. Arm lntern Med 79:528, 1973. 13. Teplitz C: Pathogenesis of Pseudomonas vasculitis and septic lesions. Arch l'athol 80:297, 1965. 14. Winn WC Jr, Myerowitz RL: The pathology of the Legionella pneumonias: a review of 74 cases and the literature. Hum Pathol 12:5, 1981. 15. Robblns SL, Cotran RS: l'athologic Basis of Disease, 2nd Ed. Philadelphia, WB Saunders Co, 1979. 16. McHenry MC, Alfidi RJ, Deodhar SD, et ah Hospital-acquired pneumonia. Med Clin North Ani 58:565, 1974. 17. Tillotson JR, Lerner A,)J: Characteristics of pneumonias caused by Bacillus proteffs. A n n Intern Med 68:287, 1968. 18. Tillotson JR, Lerner A.M.: CharacterlsUcs of pneumouias caused by Escherichiq coll. N Engl J Med 277:! 15, 1967. 19. Salomon I'F, T a m l y n T T , Grieco MH: Escherichia coli pnetnnonia: case report. Am Rev Respir Dis 102:248, 1970. 20. Katzenstein ALA, BIoor CM, Liebow A: Diffuse alveolar damage: the role of oxygen, shock, anti related factors. AmJ Pathol 85:210, 1976. 21. Atkinson BF, Pietra GG: l'ulmonary manifestations of hematologic diseases. In Fishman AP (ed): lhdmonary Diseases anti Disorders. New York, McGraw-ItiIl Book Co, 1980, 22. Katzenstein A L A , Askin FB: Interpretation and significance of pathologic findings in transbrnnchial hmg biopsy. Am J Surg Pathol 4:223, 1980. 23. Ross JS, Ellman L: Leukemic infihration of the lungs in tile chemotherapevtic era. Am J Clin Pathol 61:235, 1974. . s

,

.

.