- Email: [email protected]
Cardiac pathology in experimental brucellosis
Cardiac
Pathology
in Experimental
Brucellosis
E. Konwaler, M.D., Charlesdi. Carpenter, M.D., and Su.sumuOIlno, Ph.D., Los Angeks, Cal$.
Benjamin
INTRODIICTION
AND
BACKGROI-ND
Evidence that Brucella infection may cause some of the cardiac disease generally assumed to be of ” rheumatic” origin has been presented over the years. The evidence may be classified into four categories: (1) clinical evidence in man, (2) pathologic evidence in man, (3) dermal sensitivity to Brucellergen in patients with cardiovascular disease, and (4) cardiac lesions in experimental brucellosis. The evidence derived from the first three categories is only briefly reviewed, but the study of cardiac pathology in guinea pigs with experimental brucellosis is reported in detail. Clinical evidence for possible cardiac disease resulting from brucellosis has appeared in the literature ever since the disease was first recognized. The first account of Malta fever was published by Marstonl in 1861, and subsequently,, man)- workers have pointed to the similarity of brucellosis (undulant fever) to rheumatism and to rheumatic fever. Both Bruce,” after whom Brucella and brucellosis were named, and Hughes,3 who suggested that the disease be designated ” undulant fever,” made reference to the rheumatic nature of brucellosis. Pathologic evidence linking cardiac disease in man with brucellosis has been reported not infrequently. The literature records numerous cases of brucellosis involving the endocardium, myocardium, pericardium, cardiac valves, and smaller cardiac arteries (Carpenter and Beak”; Spink and Xelsonj; Rennie and Young6; DeGowin, Carter and Borts7; Beebe and hleneely8; Call, Baggenstoss and Merritt”). Whether the cardiac disease was the result of a primary infection with Brucella or was secondary to rheumatic heart disease was frequently questioned. (‘all, Baggenstoss and Merritty reported in detail two cases of endocarditis due to Brucella. Besides involvement of the aortic valve ill one case anti the mitral valve in the second, the authors described granulomatous lesions of the auricle and m>.ocardium. They also observed necrotizing arteritis in one case. The author’s comment : “It is interesting to note that the myocardial, renal From Center. Calif.
l,os
thca Departments Ar~gelrs. aud
t,hr
of Infectious Division
Disraws of Pathology,
and
of Pathology, University Veterans Administrat,ion
of California Hospital. Long
XIedicaI tkach.
102
KON\VAI.I’K,
(‘.\KI’I’NTI~II,
.\Yl)
.Zm. Heart
OIlNO
!nnunry.
J. 1Qhil
and vascular lesions in these cases have certain similarities to those oi rheumatic fever, subacute bacterial endocarditis due to green producing streptococci and periarteritis ~~otiosa. . Typical Xschoff bodies were not present in either of the cases and the diagnosis of active rheumatic fever ~vas not considered.” If the interpretation of the lesions in this report \vas correct, the two cases of Call, Hagg:enstoss, and IJlerrittY can be accepted as probable evidence of prirnqheart involvement in brucellosis. RecentI!-, Peer\“’ has raised the question whether rheumatic fever is the OIII!’ febrile illness which ma)- cause chronic valvular disease, and whether rheumatic fever is actually a disease entity. He states that, with the possible exception of chorea , all of the five major and seven minor manifestations of rheumatic fever set forth 1~17 Among the five major mani_ .Tonesl’ ca11 occur in brucellosis. festations of rheumatic fever, Peer). refers to carciitis espccialiy and states that, although it is 11ot generally- known, cardiac invoiveme~lt in some degree is not rare in bruceilosis. I’eery and Evans’” also believe that the electrocardiographic changes almost constantl>. observed in active brucellosis and chronic valvular disease may be a sequel of nonfatal brucellosis. ri recent study of dermai sensitivity to Brucellergen was carried out b) (‘arpenter and I,eik’” in a hospital population at t-he Long Beach \‘eterans Administration Hospital. Among 1,345 male patients tested, 385, or 28.6 per cent, reacted positivei?.. .A review of the patients exhibiting a positive reaction revealed that tlermai sensitivity. to Hrucellergen ~vas more prevalent among patients hospitalized for c:ardiovascular disease thall for other t\.ges of illness. Thus, a second group of 206 cardiac patients was simiiarl~tested, and 117, or 56.8 per cent, of the group were positive, indicating a past or present infection with Hruceila. St>.-eight cases \verc diagnosed clinically as having rheumatic heart (iisease, and 41, or 60.3 per cent, had positive skin tests. The significance of this marked prevalence of dermai sensitivit)to Bruceilergen among patients with cardiovascular disease is difficult to assess and more data are required. Yet such evidence suggests that Brucella ma\- bear some etiological relationship to morbidity from heart disease. EXPERIMENTAl..
UKCCBLLOSIS
AND
HEART
DISEASE
13ecause of the suggestive evidence of cardiac involvement in certain patients with active or chronic brucellosis, a study of cardiac disease in experimental bruceliosis in guinea pigs was undertaken. Excluding secondary infection, it may be reasonably assumed that an)- pathologic changes in the hearts of such guinea pigs were the result of infection with Bruceila. Forty guinea pigs weighing smooth cells of Brurc$la sz~is, 10 guinea pigs \vere similarly
from 250 suspended inovulated
to 300 grams were injected intraperitoneally in 0.85 per cent ph>xiologic salt solution. with Hrucrlla u.hortus strain 19 employed
vac.c.ination for Bang’s disease in c.attlc. .A third group of 10 animals c.ontrols and were normal kvith the exception of one or two cardiac. to determine the efiec.t of such trauma on the myocardium. tion.
‘I‘he guinea One guinea
pigs pip.
infected with No. 158. the
Rvucella offspring
suis were sacrificed of a mother infected
with 3 s 107 In addition, iu calfhood
nzs observed as uninfected punctures prior to autopq
from 39 to 383 with Brur~lla
days suis
after inoculawas swrificed
6 days after birth. Prior to postmortem examination, an intradermal test with 0.2 ml. Hrucellergell was carried out. .At autopsy, bacteriologic and serologic examinations were made to determine the presence of infection. The heart, together with portions of the lung. liver, spleen, kidneys, adrenal and any other grossly abnormal tissues were removed and tixrd ill 10 per cent formalill. ;\fter fixation, sections from the heart were made at three levels, i.e., at the junctiotl of thegrcat \-essels, at the nIIriculo-~entricul~tr junction, and at the mid-\ entric-ular area. Slides were prepared from the cardiac tissue, stained with hematoxylin and eosin, and examined microscopicall>-. Similar observations were made on the normal guinea pigs and on those infected xvith Rrucclla clhoutusPstrain 19. This report includes pathologic studies on 37 guinea pigs, 32 inoculated with Rurrlln wis and 5 with Rr~uzella abortus -straill 19.
Fig. I.-Sulxndocardial nucleated giant cell is seen. 2 and 3 for cellular detail.
lesion at Hematosylin
A. consisting and rosin
of a localized collection of ~11s. st,ain; magnification X 100, reducwl
AL
IS, a mu1t.i.. ?i. SW Wigs.
The majority of the guinea pigs infected with b’v. suis showed onl!. minimal clinical evidence of infection. Some developed transient arthritis of either the tarsal or carpal joints and failed to gain weight as rapidly as did the control group of normal guinea pigs. A few experienced an acute illness accompanied by loss of weight and ruffled fur. In several instances, ophthalmia and neurologic signs of disease developed. Three guinea pigs, SOS. 11, 25, and 37, died with brucellosis, 58, 65, and 242 days, respectively, after inoculation. With the esception of these three, all of the guinea pigs inoculated with Br. suis showed
104
KON\VALI;K,
,^
CAWENTISK,
,-
ANI)
OHNO
Fig.
4.
Fig.
5.
Wig. 4.-Subendocardial lesion showing fibrinoid nncrosis with cellular reaction consisting of fibrocytes, macrophages, occasional lymphocyw, and a few polymorphonurlear lrukocytes. A resemblance to the lesion of acute rheumatic endocarditis may he observed. Hematoxylin and eosin stain; magnification X300, reduced v4, Fig. 5.-Epicardial. perivascular granulomatous lesion consisting chiefly of fixed iissw cells and lymphoid cells. Hematoxylin and eosin stain: magnification X300, reduced I<.
evidence of infection at the time of autopsy, as indicated by either the isolation of the organism from the blood of the spleen or heart or by the demonstration of specific agglutinins. The guinea pigs evidence of disease. either agglutinins or as controls remained
infected with BY. abortzss-strain 19 showed little clinical Of the 10 pigs inoculated with Br. abortus, 3 failed to show organisms in culture. The uninfected guinea pigs maintained well.
Gross Z’nthologir Findings.~Macroscopicalli~,, the guinea pigs infected with BY. suis showed typical granulomatous lesions of experimental brucellosis in the spleen, I>mph nodes, liver, joints, and epididymides, as described in the literature b\, Fab>xll’4 and 1~~. Hraude.Ij The gross appearance of the heart, how'I‘ARLK I. INFISCTEJ)
BACTERIOLOGIC \VITH BRIWELLA
AND Sues
SEROLOGIC OB~ER~~T~ON~ ox AND SUBSEQURNTLY EUMIN~~
NllMBE:R
HRuCI;LLA
AGGLVTININ TITER
GUNEA
PIGS ESPBRIMENTALL~ FOR CARDIOVASCULAR DISEASE
SI'LEEN CULTL:RE FOR BRUCELLA
8. 9. 11.
116 2.57
l--l 0240 I-
2560
13. 15. 16.
249 2.57
Il1
2560 2560 2560
Segative Segative
I--
2560
Positive
CL 163
17. 2s. 26. 27. 32. .34. 35. 36. 37. 40. 42. 4.z. 44. 4.5. 46. 49. 54. 58. 62. 63. 67. 76. 8.7. 87. 88. 98. 12.z.*
246 58 383
1-m 5120 l--- 5120
125 205
f’ositive Positive f’osi tive Positive Positive Negative
l-10240 l2560
1.53 102 97 65
lp
5120
1
2.560
lm ll--lp
1280 S120 2560 2S60
Positive Positive
136 76 82 96
136 136 107 120 178 77 98 100 46 64 253 2 .io 98 64 70 39 70
141.* 142."
143.* 158.t *Inoculaterl t5lotherinfected
Segative Positive
with
Urucella abortus. with Brucdla suis.
l-10240 IL10210 I-- 5120
Positive Positive Positive
l-- 2560 l- 1280 l- 2560
Positive Positive
Illp lll1~ 1
llm
Examined
Positive
6 days
5120 5120 160 2560 1280 80
.Uegative Negative Positive Negative Positive Negative Negative Negative Negative Negative
40
160 80 640
after
birth.
ever, was normal. To our knowledge, gross involvement of the heart has 1101 been described in experimental brucellosis. The guinea pigs inoculated with BY. abortus-strain 19 showed few or no lesions, and the group of normal controls was free from gross evidence of disease. Jficroscopic Pathologic Findings.--Of the 32 guinea pigs infected with Br. suis whose hearts were examined, 1.5, or 46.X per cent, showed microscopic lesions which were considered to be the result of the experimental infection. Four other animals infected with Br. sz~is had questionable small increases in Anitschkon myocytes. No lesions were observed in the hearts of the remaining 17. The hearts of the normal guinea pigs and those inoculated with BY. clbortus-strain 19 likcwise revealed no pathologic changes.
Fig. 6:-Focal
areas
of interst,itial cell. Hmmt,oxylin
myocarditis and eosin
with fixed tissue cell stain; magnification
proliferation and occasional x700, r~~~ucrrl li.
lyrnphoid
The lesions were predominantl\. located either in the endocardium or in the epicardium, with a smaller number in the m>.ocardium. Generally, the lesions were localized perivascularly. Only one guinea pig, A’o. 8, showed evidence of a necrotizing lesion of a vessel wall. The lesions, in general, may be classified into three groups. Group 1 included inflammatory lesions in the epicardium which consisted of collections of lq-mphocytes, plasma cells, and rare polgmorphonuclear leukocytes in an edematous
108
KONWALER,
CARPENTER,
AND
OHNO
.\m. Hmrt J. January, 1960
connective tissue stroma. Eosinophils were not seen (Figs. 1,B and 3j. Group 2 (Figs. l,A, 2, 4, 5) included granulomatous lesions, frequentI>subendocardial but also located in the epicardium and, occasionally, in the tn>.ocardium. These granulomatous lesions generally consisted of fixed tissue or mesenchymal cells which varied in size and shape and could probably be classified as macrophages. Small areas of necrosis were occasionally seen in the center of such lesions, but typical fibrinoid necrosis was not seen. Some of the lesions also c-ontairted inflammatory cells, mainly lymphocytes and polymorphonuclear leukocytes. Such lesions resemble some of the changes seen in acute rheumatic endocarditis. (;roup 3 (Fig. 3) included areas of apparent interstitial m>.ocarditis, with proliferation of fixed tissue cells attd occasional Ih.mphoid cells. In all sections the increase in Anitschkow c-ells was far more promittettt than in normal controls. While occasionall\. such areas resembled an d%schoff body, an unequivocal Aschoff type of granuloma was not observed (Fig. 3). The hearts in this series showed no valvular involvetnent. (However, in one oi the animals, No. 24, in a later series, a granulomatous lesion was presettt at the base of the mitral valve.) It is significant that the lesions are not numerous in any one section and are not so diffusely distributed as in rheumatic fever. In general, one or two lesions are seen in a section.
The production of lesions by an experimental infection in animals must be interpreted with considerable care inasmuch as intercurrent infection rna)~ develop during an observation period. When inoculated animals develop specific evidence of disease, however, as proved b>. serologic and bacteriologic examination, and when the lesions produced are far more numerous and vary from a nonspecific lesion, one may be justified in assuming that such lesions resulted from the experimental inoculation. In experimental brucellosis, cardiac lesions were produced in a significant number of guinea pigs inoculated with Bruce&z suis but not in those inoculated with Brucella abortus-strain 19. In the literature, reference is rarely made to cardiac involvement in experimental brucellosis in guinea pigs. Only Fablrattt” made reference to examination of the heart and, in photomicrographs, showed a perivascular lesion in the epicardium. Braude,t5 ott the other hand, in ;III estensive study of experimental hrucellosis, made no reierence whatever to examination of the hearts. Most investigators have examined microscopicall>. o~tl\r those tissues which show macroscopic lesions, and because the heart, in experimental brucellosis, reveals no gross abnormalit.ies, the cardiovascular Itathologl, has received little attention. Whether cardiac lesions similar to those of rheumatic fever can be produced in experimental brucellosis depends upon interpretation of the lesions. In our opinion, proliferation of Anitschkow myocytes does not necessarily constitute formation of an Aschoff body. Large or multinuclear cells such as those seen in Aschoff bodies were not observed in the experimental cardiac lesions. Furthermore, fibrinoid degeneration was not noted. On the other hand, many of the
Volunle Number
59 1
i‘ARDIhC
PATiXOLOGH
IN
EXPERIMENTAL
109
BRUCELLOSIS
lesions were not unlike the nonspecific endocardia1, epicardial, and myocardial lesions consisting of cells resembling macrophages with some circulating inflammatory cells. RichI described five cardiac lesions characteristic of rheumatic fever: (1) focal alterations in the collagen of the connective tissue, (2) the presence of the Aschoff body-, (3) focal and diffuse inflammatory lesions, (4) focal alterations in cardiac muscle, and (5) verrucous valvular lesions. In our series thus far, three of the above criteria, namely, focal alterations in the collagen, focal infammatory lesions, and focal alterations in the cardiac muscle, were observed. Furthermore, in a recent animal a granulomatous lesion at the base of the mitral valve was observed. The more specific lesions of rheumatic fever, namely, Aschoff bodies and verrucous valvular lesions, were not definitely established after a single inoculation of Brucella. Certain lesions may be referred to as Aschoff-like bodies, however. The fact that typical Aschoff bodies were not observed may be due to the experimental host’s reaction to the infection, which might be at variance with the reaction in man. Studies employing multiple injections to determine the ty.pe of lesions which may develop are in progress. Thus, these findings suggest that in human beings, brucellosis may cause certain lesions not unlike those present in rheumatic fever. 111 a recent report, Saphir17 reviewed the morphologic filldings in what he considers a true Aschoff body and stated, “Strictest criteria of what constitutes an Aschoff body are of utmost importance, not only from the morphologic point of view to recognize true rheumatic heart disease but especially for experimental and etiologic reasons. . . As stated, it is wiser to err and call a lesion ‘doubtful’ and not make a definite and specific diagnosis than to designate lesions somewhat resembling Aschoff bodies as true Aschoff bodies.”
(‘ardiac pathology in 37 guinea pigs was studied at intervals of from 39 to 3X3 days after a single inoculation with either Brucella suis or Brucellu abortusstrain 19. Although definite microscopic cardiac lesions were present in 15, or 46.8 per cent, of the group infected with Brucella suis, they tended to resemble the so-called nonspecific lesions observed after injection with dead streptococci, foreign proteins, dysentery bacilli, serum, and a number of other substances. So lesions completely identical to Aschoff bodies were observed. The experimental evidence indicates that Brucella infection causes microscopic lesions in the heart of the guinea pig, and that the heart is not as resistant to the infection as was inferred from gross examination. The Konwaler Institute
1. 2. 3.
authors are grateful to Kathryn M. Miller for technical assistance and to Ruth H. for preparation of the microscopic sections. Ik. \Y. C. Manion of the Armed Forces of Pathology kindly reviewed the sections from several of the cases herein reported.
hlarston,
J. A.: Quoted by Bloomlield, A. L.: Bibliography on Brucellosis, 3208, 19.56. Bruce, D.: Observations on Malta Fever, Brit. M. J. l:llOl, 1889. Hughes, L. M.: Mediterranean, Malta or Undulant Fever, New York, 1897, c-0.
J. Chron.
T)is.
Macmillan
and
110 1. 5. 6. 7. 8. 9. 10. Il. 12. 13.
11. 15. 16. 17.
KONWALEK,
(:AKWNTI,K,
.'iND
OIINO
.\nl. Heart J. January, 1960
Carpenter, C. AI., and Beak, K. X.: Agglutinins for Brucella ribortus in Blood of Man (Includes Case Report of Subacute Bacterial Endocarditis), J. Infect. Dis. 39:220, 1926. Spink, \V. IV., and ANelson, ;\. .I.: Brucella Endocarditis --Case Report, Ann. Int. Med. 13:721, 1939. Rennie. I. Ii.. and Youw. C. I.: Malignant Elldocarditis I)ue to Brucella Abortus. Brit. $1: J. 11412, 1936. -’ _I “ L)eGowin, E. L., Carter, J. Ii., and Borts, I. I-1.: Brucella Endocarditis, Case Report, .U. HEART J. 30:77, 1915. Beebe, R. T., and Rleneely, J. K.: Bruce& ~lelitensis Endocarditis, AM. HEART J. 38:788, 1949. Call, J. D., Baggenstoss, A. H., and Merritt, LV. A.: Endocarditis Ijue to Brucella--Report of 2 Cases, Am. J. Clin. Path. 14508, 1944. I’eeq,, T. RI.: Brucellosis and Heart Disease, Postgrad. M.J. 19:323, 1956. Jones, ‘I’. L).: The Diagnosis of Rheumatic Fever, J.A.M.A. 126:481, 1944. Peery, 7‘. LI., and E\-ans, J. M.: Brucellosis and Heart Disease. III. Chronic Valvular Heart Disease Following Nonfatal Brucellosis, Ann. Int. Med. 49568, 19.58. Carpenter, C. &I., and Leik, D. W.: ‘The Prevalence of Dermal Sensitivity to Brucellergen .Xmong Patients \f?th Cardiovascular Disease. (To be published.) Presented before the American Society of Tropical Medicine and Hygiene, November 1, 1956, New Orleans, La. Fabyan, RI.: ;\ Contribution to Pathogen&s of Brucella Abortus, J. Med. Kes. 26:441, 1912. Braude, A. I.: Studies in the Pathology and I’athogenesis of Esperimental Brucellosis, J. Infect. Dis. 89:76, 1951. Rich, A. R., and Gregory, J. E.: Experimental Evidence That Lesions With the Basic Characteristics of Rhemnatic Carditis Can Result From Xnaphylactic Hypersensitivity, Bull. Johns Hopkins Hosp. 73:23?, 1943. Saphir, 0.: The -kchoff Nodule (:\n Editorial), :\m. J. Clili. f’ath. 31534, 1959.
Pathology
in Experimental
Brucellosis
E. Konwaler, M.D., Charlesdi. Carpenter, M.D., and Su.sumuOIlno, Ph.D., Los Angeks, Cal$.
Benjamin
INTRODIICTION
AND
BACKGROI-ND
Evidence that Brucella infection may cause some of the cardiac disease generally assumed to be of ” rheumatic” origin has been presented over the years. The evidence may be classified into four categories: (1) clinical evidence in man, (2) pathologic evidence in man, (3) dermal sensitivity to Brucellergen in patients with cardiovascular disease, and (4) cardiac lesions in experimental brucellosis. The evidence derived from the first three categories is only briefly reviewed, but the study of cardiac pathology in guinea pigs with experimental brucellosis is reported in detail. Clinical evidence for possible cardiac disease resulting from brucellosis has appeared in the literature ever since the disease was first recognized. The first account of Malta fever was published by Marstonl in 1861, and subsequently,, man)- workers have pointed to the similarity of brucellosis (undulant fever) to rheumatism and to rheumatic fever. Both Bruce,” after whom Brucella and brucellosis were named, and Hughes,3 who suggested that the disease be designated ” undulant fever,” made reference to the rheumatic nature of brucellosis. Pathologic evidence linking cardiac disease in man with brucellosis has been reported not infrequently. The literature records numerous cases of brucellosis involving the endocardium, myocardium, pericardium, cardiac valves, and smaller cardiac arteries (Carpenter and Beak”; Spink and Xelsonj; Rennie and Young6; DeGowin, Carter and Borts7; Beebe and hleneely8; Call, Baggenstoss and Merritt”). Whether the cardiac disease was the result of a primary infection with Brucella or was secondary to rheumatic heart disease was frequently questioned. (‘all, Baggenstoss and Merritty reported in detail two cases of endocarditis due to Brucella. Besides involvement of the aortic valve ill one case anti the mitral valve in the second, the authors described granulomatous lesions of the auricle and m>.ocardium. They also observed necrotizing arteritis in one case. The author’s comment : “It is interesting to note that the myocardial, renal From Center. Calif.
l,os
thca Departments Ar~gelrs. aud
t,hr
of Infectious Division
Disraws of Pathology,
and
of Pathology, University Veterans Administrat,ion
of California Hospital. Long
XIedicaI tkach.
102
KON\VAI.I’K,
(‘.\KI’I’NTI~II,
.\Yl)
.Zm. Heart
OIlNO
!nnunry.
J. 1Qhil
and vascular lesions in these cases have certain similarities to those oi rheumatic fever, subacute bacterial endocarditis due to green producing streptococci and periarteritis ~~otiosa. . Typical Xschoff bodies were not present in either of the cases and the diagnosis of active rheumatic fever ~vas not considered.” If the interpretation of the lesions in this report \vas correct, the two cases of Call, Hagg:enstoss, and IJlerrittY can be accepted as probable evidence of prirnqheart involvement in brucellosis. RecentI!-, Peer\“’ has raised the question whether rheumatic fever is the OIII!’ febrile illness which ma)- cause chronic valvular disease, and whether rheumatic fever is actually a disease entity. He states that, with the possible exception of chorea , all of the five major and seven minor manifestations of rheumatic fever set forth 1~17 Among the five major mani_ .Tonesl’ ca11 occur in brucellosis. festations of rheumatic fever, Peer). refers to carciitis espccialiy and states that, although it is 11ot generally- known, cardiac invoiveme~lt in some degree is not rare in bruceilosis. I’eery and Evans’” also believe that the electrocardiographic changes almost constantl>. observed in active brucellosis and chronic valvular disease may be a sequel of nonfatal brucellosis. ri recent study of dermai sensitivity to Brucellergen was carried out b) (‘arpenter and I,eik’” in a hospital population at t-he Long Beach \‘eterans Administration Hospital. Among 1,345 male patients tested, 385, or 28.6 per cent, reacted positivei?.. .A review of the patients exhibiting a positive reaction revealed that tlermai sensitivity. to Hrucellergen ~vas more prevalent among patients hospitalized for c:ardiovascular disease thall for other t\.ges of illness. Thus, a second group of 206 cardiac patients was simiiarl~tested, and 117, or 56.8 per cent, of the group were positive, indicating a past or present infection with Hruceila. St>.-eight cases \verc diagnosed clinically as having rheumatic heart (iisease, and 41, or 60.3 per cent, had positive skin tests. The significance of this marked prevalence of dermai sensitivit)to Bruceilergen among patients with cardiovascular disease is difficult to assess and more data are required. Yet such evidence suggests that Brucella ma\- bear some etiological relationship to morbidity from heart disease. EXPERIMENTAl..
UKCCBLLOSIS
AND
HEART
DISEASE
13ecause of the suggestive evidence of cardiac involvement in certain patients with active or chronic brucellosis, a study of cardiac disease in experimental bruceliosis in guinea pigs was undertaken. Excluding secondary infection, it may be reasonably assumed that an)- pathologic changes in the hearts of such guinea pigs were the result of infection with Bruceila. Forty guinea pigs weighing smooth cells of Brurc$la sz~is, 10 guinea pigs \vere similarly
from 250 suspended inovulated
to 300 grams were injected intraperitoneally in 0.85 per cent ph>xiologic salt solution. with Hrucrlla u.hortus strain 19 employed
vac.c.ination for Bang’s disease in c.attlc. .A third group of 10 animals c.ontrols and were normal kvith the exception of one or two cardiac. to determine the efiec.t of such trauma on the myocardium. tion.
‘I‘he guinea One guinea
pigs pip.
infected with No. 158. the
Rvucella offspring
suis were sacrificed of a mother infected
with 3 s 107 In addition, iu calfhood
nzs observed as uninfected punctures prior to autopq
from 39 to 383 with Brur~lla
days suis
after inoculawas swrificed
6 days after birth. Prior to postmortem examination, an intradermal test with 0.2 ml. Hrucellergell was carried out. .At autopsy, bacteriologic and serologic examinations were made to determine the presence of infection. The heart, together with portions of the lung. liver, spleen, kidneys, adrenal and any other grossly abnormal tissues were removed and tixrd ill 10 per cent formalill. ;\fter fixation, sections from the heart were made at three levels, i.e., at the junctiotl of thegrcat \-essels, at the nIIriculo-~entricul~tr junction, and at the mid-\ entric-ular area. Slides were prepared from the cardiac tissue, stained with hematoxylin and eosin, and examined microscopicall>-. Similar observations were made on the normal guinea pigs and on those infected xvith Rrucclla clhoutusPstrain 19. This report includes pathologic studies on 37 guinea pigs, 32 inoculated with Rurrlln wis and 5 with Rr~uzella abortus -straill 19.
Fig. I.-Sulxndocardial nucleated giant cell is seen. 2 and 3 for cellular detail.
lesion at Hematosylin
A. consisting and rosin
of a localized collection of ~11s. st,ain; magnification X 100, reducwl
AL
IS, a mu1t.i.. ?i. SW Wigs.
The majority of the guinea pigs infected with b’v. suis showed onl!. minimal clinical evidence of infection. Some developed transient arthritis of either the tarsal or carpal joints and failed to gain weight as rapidly as did the control group of normal guinea pigs. A few experienced an acute illness accompanied by loss of weight and ruffled fur. In several instances, ophthalmia and neurologic signs of disease developed. Three guinea pigs, SOS. 11, 25, and 37, died with brucellosis, 58, 65, and 242 days, respectively, after inoculation. With the esception of these three, all of the guinea pigs inoculated with Br. suis showed
104
KON\VALI;K,
,^
CAWENTISK,
,-
ANI)
OHNO
Fig.
4.
Fig.
5.
Wig. 4.-Subendocardial lesion showing fibrinoid nncrosis with cellular reaction consisting of fibrocytes, macrophages, occasional lymphocyw, and a few polymorphonurlear lrukocytes. A resemblance to the lesion of acute rheumatic endocarditis may he observed. Hematoxylin and eosin stain; magnification X300, reduced v4, Fig. 5.-Epicardial. perivascular granulomatous lesion consisting chiefly of fixed iissw cells and lymphoid cells. Hematoxylin and eosin stain: magnification X300, reduced I<.
evidence of infection at the time of autopsy, as indicated by either the isolation of the organism from the blood of the spleen or heart or by the demonstration of specific agglutinins. The guinea pigs evidence of disease. either agglutinins or as controls remained
infected with BY. abortzss-strain 19 showed little clinical Of the 10 pigs inoculated with Br. abortus, 3 failed to show organisms in culture. The uninfected guinea pigs maintained well.
Gross Z’nthologir Findings.~Macroscopicalli~,, the guinea pigs infected with BY. suis showed typical granulomatous lesions of experimental brucellosis in the spleen, I>mph nodes, liver, joints, and epididymides, as described in the literature b\, Fab>xll’4 and 1~~. Hraude.Ij The gross appearance of the heart, how'I‘ARLK I. INFISCTEJ)
BACTERIOLOGIC \VITH BRIWELLA
AND Sues
SEROLOGIC OB~ER~~T~ON~ ox AND SUBSEQURNTLY EUMIN~~
NllMBE:R
HRuCI;LLA
AGGLVTININ TITER
GUNEA
PIGS ESPBRIMENTALL~ FOR CARDIOVASCULAR DISEASE
SI'LEEN CULTL:RE FOR BRUCELLA
8. 9. 11.
116 2.57
l--l 0240 I-
2560
13. 15. 16.
249 2.57
Il1
2560 2560 2560
Segative Segative
I--
2560
Positive
CL 163
17. 2s. 26. 27. 32. .34. 35. 36. 37. 40. 42. 4.z. 44. 4.5. 46. 49. 54. 58. 62. 63. 67. 76. 8.7. 87. 88. 98. 12.z.*
246 58 383
1-m 5120 l--- 5120
125 205
f’ositive Positive f’osi tive Positive Positive Negative
l-10240 l2560
1.53 102 97 65
lp
5120
1
2.560
lm ll--lp
1280 S120 2560 2S60
Positive Positive
136 76 82 96
136 136 107 120 178 77 98 100 46 64 253 2 .io 98 64 70 39 70
141.* 142."
143.* 158.t *Inoculaterl t5lotherinfected
Segative Positive
with
Urucella abortus. with Brucdla suis.
l-10240 IL10210 I-- 5120
Positive Positive Positive
l-- 2560 l- 1280 l- 2560
Positive Positive
Illp lll1~ 1
llm
Examined
Positive
6 days
5120 5120 160 2560 1280 80
.Uegative Negative Positive Negative Positive Negative Negative Negative Negative Negative
40
160 80 640
after
birth.
ever, was normal. To our knowledge, gross involvement of the heart has 1101 been described in experimental brucellosis. The guinea pigs inoculated with BY. abortus-strain 19 showed few or no lesions, and the group of normal controls was free from gross evidence of disease. Jficroscopic Pathologic Findings.--Of the 32 guinea pigs infected with Br. suis whose hearts were examined, 1.5, or 46.X per cent, showed microscopic lesions which were considered to be the result of the experimental infection. Four other animals infected with Br. sz~is had questionable small increases in Anitschkon myocytes. No lesions were observed in the hearts of the remaining 17. The hearts of the normal guinea pigs and those inoculated with BY. clbortus-strain 19 likcwise revealed no pathologic changes.
Fig. 6:-Focal
areas
of interst,itial cell. Hmmt,oxylin
myocarditis and eosin
with fixed tissue cell stain; magnification
proliferation and occasional x700, r~~~ucrrl li.
lyrnphoid
The lesions were predominantl\. located either in the endocardium or in the epicardium, with a smaller number in the m>.ocardium. Generally, the lesions were localized perivascularly. Only one guinea pig, A’o. 8, showed evidence of a necrotizing lesion of a vessel wall. The lesions, in general, may be classified into three groups. Group 1 included inflammatory lesions in the epicardium which consisted of collections of lq-mphocytes, plasma cells, and rare polgmorphonuclear leukocytes in an edematous
108
KONWALER,
CARPENTER,
AND
OHNO
.\m. Hmrt J. January, 1960
connective tissue stroma. Eosinophils were not seen (Figs. 1,B and 3j. Group 2 (Figs. l,A, 2, 4, 5) included granulomatous lesions, frequentI>subendocardial but also located in the epicardium and, occasionally, in the tn>.ocardium. These granulomatous lesions generally consisted of fixed tissue or mesenchymal cells which varied in size and shape and could probably be classified as macrophages. Small areas of necrosis were occasionally seen in the center of such lesions, but typical fibrinoid necrosis was not seen. Some of the lesions also c-ontairted inflammatory cells, mainly lymphocytes and polymorphonuclear leukocytes. Such lesions resemble some of the changes seen in acute rheumatic endocarditis. (;roup 3 (Fig. 3) included areas of apparent interstitial m>.ocarditis, with proliferation of fixed tissue cells attd occasional Ih.mphoid cells. In all sections the increase in Anitschkow c-ells was far more promittettt than in normal controls. While occasionall\. such areas resembled an d%schoff body, an unequivocal Aschoff type of granuloma was not observed (Fig. 3). The hearts in this series showed no valvular involvetnent. (However, in one oi the animals, No. 24, in a later series, a granulomatous lesion was presettt at the base of the mitral valve.) It is significant that the lesions are not numerous in any one section and are not so diffusely distributed as in rheumatic fever. In general, one or two lesions are seen in a section.
The production of lesions by an experimental infection in animals must be interpreted with considerable care inasmuch as intercurrent infection rna)~ develop during an observation period. When inoculated animals develop specific evidence of disease, however, as proved b>. serologic and bacteriologic examination, and when the lesions produced are far more numerous and vary from a nonspecific lesion, one may be justified in assuming that such lesions resulted from the experimental inoculation. In experimental brucellosis, cardiac lesions were produced in a significant number of guinea pigs inoculated with Bruce&z suis but not in those inoculated with Brucella abortus-strain 19. In the literature, reference is rarely made to cardiac involvement in experimental brucellosis in guinea pigs. Only Fablrattt” made reference to examination of the heart and, in photomicrographs, showed a perivascular lesion in the epicardium. Braude,t5 ott the other hand, in ;III estensive study of experimental hrucellosis, made no reierence whatever to examination of the hearts. Most investigators have examined microscopicall>. o~tl\r those tissues which show macroscopic lesions, and because the heart, in experimental brucellosis, reveals no gross abnormalit.ies, the cardiovascular Itathologl, has received little attention. Whether cardiac lesions similar to those of rheumatic fever can be produced in experimental brucellosis depends upon interpretation of the lesions. In our opinion, proliferation of Anitschkow myocytes does not necessarily constitute formation of an Aschoff body. Large or multinuclear cells such as those seen in Aschoff bodies were not observed in the experimental cardiac lesions. Furthermore, fibrinoid degeneration was not noted. On the other hand, many of the
Volunle Number
59 1
i‘ARDIhC
PATiXOLOGH
IN
EXPERIMENTAL
109
BRUCELLOSIS
lesions were not unlike the nonspecific endocardia1, epicardial, and myocardial lesions consisting of cells resembling macrophages with some circulating inflammatory cells. RichI described five cardiac lesions characteristic of rheumatic fever: (1) focal alterations in the collagen of the connective tissue, (2) the presence of the Aschoff body-, (3) focal and diffuse inflammatory lesions, (4) focal alterations in cardiac muscle, and (5) verrucous valvular lesions. In our series thus far, three of the above criteria, namely, focal alterations in the collagen, focal infammatory lesions, and focal alterations in the cardiac muscle, were observed. Furthermore, in a recent animal a granulomatous lesion at the base of the mitral valve was observed. The more specific lesions of rheumatic fever, namely, Aschoff bodies and verrucous valvular lesions, were not definitely established after a single inoculation of Brucella. Certain lesions may be referred to as Aschoff-like bodies, however. The fact that typical Aschoff bodies were not observed may be due to the experimental host’s reaction to the infection, which might be at variance with the reaction in man. Studies employing multiple injections to determine the ty.pe of lesions which may develop are in progress. Thus, these findings suggest that in human beings, brucellosis may cause certain lesions not unlike those present in rheumatic fever. 111 a recent report, Saphir17 reviewed the morphologic filldings in what he considers a true Aschoff body and stated, “Strictest criteria of what constitutes an Aschoff body are of utmost importance, not only from the morphologic point of view to recognize true rheumatic heart disease but especially for experimental and etiologic reasons. . . As stated, it is wiser to err and call a lesion ‘doubtful’ and not make a definite and specific diagnosis than to designate lesions somewhat resembling Aschoff bodies as true Aschoff bodies.”
(‘ardiac pathology in 37 guinea pigs was studied at intervals of from 39 to 3X3 days after a single inoculation with either Brucella suis or Brucellu abortusstrain 19. Although definite microscopic cardiac lesions were present in 15, or 46.8 per cent, of the group infected with Brucella suis, they tended to resemble the so-called nonspecific lesions observed after injection with dead streptococci, foreign proteins, dysentery bacilli, serum, and a number of other substances. So lesions completely identical to Aschoff bodies were observed. The experimental evidence indicates that Brucella infection causes microscopic lesions in the heart of the guinea pig, and that the heart is not as resistant to the infection as was inferred from gross examination. The Konwaler Institute
1. 2. 3.
authors are grateful to Kathryn M. Miller for technical assistance and to Ruth H. for preparation of the microscopic sections. Ik. \Y. C. Manion of the Armed Forces of Pathology kindly reviewed the sections from several of the cases herein reported.
hlarston,
J. A.: Quoted by Bloomlield, A. L.: Bibliography on Brucellosis, 3208, 19.56. Bruce, D.: Observations on Malta Fever, Brit. M. J. l:llOl, 1889. Hughes, L. M.: Mediterranean, Malta or Undulant Fever, New York, 1897, c-0.
J. Chron.
T)is.
Macmillan
and
110 1. 5. 6. 7. 8. 9. 10. Il. 12. 13.
11. 15. 16. 17.
KONWALEK,
(:AKWNTI,K,
.'iND
OIINO
.\nl. Heart J. January, 1960
Carpenter, C. AI., and Beak, K. X.: Agglutinins for Brucella ribortus in Blood of Man (Includes Case Report of Subacute Bacterial Endocarditis), J. Infect. Dis. 39:220, 1926. Spink, \V. IV., and ANelson, ;\. .I.: Brucella Endocarditis --Case Report, Ann. Int. Med. 13:721, 1939. Rennie. I. Ii.. and Youw. C. I.: Malignant Elldocarditis I)ue to Brucella Abortus. Brit. $1: J. 11412, 1936. -’ _I “ L)eGowin, E. L., Carter, J. Ii., and Borts, I. I-1.: Brucella Endocarditis, Case Report, .U. HEART J. 30:77, 1915. Beebe, R. T., and Rleneely, J. K.: Bruce& ~lelitensis Endocarditis, AM. HEART J. 38:788, 1949. Call, J. D., Baggenstoss, A. H., and Merritt, LV. A.: Endocarditis Ijue to Brucella--Report of 2 Cases, Am. J. Clin. Path. 14508, 1944. I’eeq,, T. RI.: Brucellosis and Heart Disease, Postgrad. M.J. 19:323, 1956. Jones, ‘I’. L).: The Diagnosis of Rheumatic Fever, J.A.M.A. 126:481, 1944. Peery, 7‘. LI., and E\-ans, J. M.: Brucellosis and Heart Disease. III. Chronic Valvular Heart Disease Following Nonfatal Brucellosis, Ann. Int. Med. 49568, 19.58. Carpenter, C. &I., and Leik, D. W.: ‘The Prevalence of Dermal Sensitivity to Brucellergen .Xmong Patients \f?th Cardiovascular Disease. (To be published.) Presented before the American Society of Tropical Medicine and Hygiene, November 1, 1956, New Orleans, La. Fabyan, RI.: ;\ Contribution to Pathogen&s of Brucella Abortus, J. Med. Kes. 26:441, 1912. Braude, A. I.: Studies in the Pathology and I’athogenesis of Esperimental Brucellosis, J. Infect. Dis. 89:76, 1951. Rich, A. R., and Gregory, J. E.: Experimental Evidence That Lesions With the Basic Characteristics of Rhemnatic Carditis Can Result From Xnaphylactic Hypersensitivity, Bull. Johns Hopkins Hosp. 73:23?, 1943. Saphir, 0.: The -kchoff Nodule (:\n Editorial), :\m. J. Clili. f’ath. 31534, 1959.