Heart and Vessels (1986) 2:55-59 Heart "° Vessels

© Springer-Verlag 1986

Rickettsial perimyocarditis A follow-up study* Bernhard Maisch

Medizinische Universitgtsklinik, Josef-Schneider-Str. 2, D-8700 Wfirzburg, Federal Republic of Germany

Summary. Sera and lymphocytes from a 37-year-old male patient with acute perimyocarditis during a Q-fever endemic were analyzed for antibody and cell- mediated immune reactions and followed up 28 months later. Circulating autoantibodies against myocardial tissue were assessed by indirect immuno- fluorescence. Cytolysis of vital contracting rat car- diocytes, by antimyolemmal antibodies and comple- ment, and lymphocytotoxicity, with and without the patient's serum, were evaluated and compared with the results obtained in ten patients suffering from Q-fever without perimyocardial involvement and with 40 healthy" subjects.

Antimyolemmal antibodies (AMLA), a muscle- specific subtype of antisarcolemmal antibodies, were demonstrated by immunofluorescence in the one pa- tient with Q-fever perimyocarditis in titers of up to 1 : 320 but not in the controls. AMLA induced cytoly- sis of myocytes in the presence of complement. Both AMLA and cytolytic serum activity could be ab- sorbed in all sera of this patient by using Coxiella burnetii. Only marginal lymphocytotoxicity against heterologous cardiocytes was detected in the early phase and again during the follow-up 2 years later in the Q-fever myocarditis patient but not in any of the noncarditic Q-fever cases nor in controls. It is postu- lated that cross-reacting, complement-fixing, cytolytic autoantibodies against the cardiac myolemma are operative either as a cause of cardiac damage or a consequence, pointing to a secondary immunopatho- genesis of chronic Q-fever perimyocarditis.

Key words: Q-fever - Myocarditis - Immunology Antisarcolemmal antibodies ...... Antimyolemmal anti- bodies - Cytolysis of heart cells

* supported by grants of the Deutsche Forschungsgemeinschaft (Ma 780/1-5)

Viral and bacterial diseases involving the peri- and myocardium are often associated with antisarcolem- real or antimyolemmal and non-organ-specific anti- endothelial antibodies [1-4]. During or after ricket- tsial infections, myocarditis [5-7] and endocarditis [8-10] have been described but occur rarely. We have therefore followed up a 37-year-old patient with acute perimyocardial disease due_ to, or in association with, an infection by Rickettsia burnetii to find out if a secondary immunopathogenesis may be involved as proposed for rheumatic carditis [11].

Material and methods

The indirect immunofluorescence technique [12] was performed on cryostat sections of homologous, operatively resected myocardium from patients with congenital heart disease of bloodgroup 0; the technique has been described in detail previously [1-3]. Antibodies directed against the glycocalix or myolemma were tested on isolated intact rat myocytes (adult type; isolation according to Powell and Twist and Maisch et al. [13, 14]) and on isolated intact human atrial myocytes from intraoperatively resected atria, which were minced and treated consecutively with crude collagenase incubations until vital heart cells were obtained (unpublished data). By treatment with collagenase during the perfusion and isolation precedure, the isolated myocytes were deprived of the collagenous perimysium (15). Assays of cardiocytotoxicity were carried out as already described in detail elsewhere [2, 3]. The following assays were performed.

Antibody.mediated cytolysis ( AMC; cytolytic serum activity in the presence of complement)

The index of cytolysis was calculated from the half-life of cardio- cytes in the presence of the patient's serum and complement (tsoi) when compared with the half-life of cardiocytes in the control pool serum and complement (ts0c), according to the formula:

Cla~tc _ is01; normals (n = 70) Clam c = 0.92 + 0.07 lsoc

An effective cytolysis was present when the C l a ~ c < 0.78 (mean - 2 SD). Corresponding assays of hepatocytolysis were carried out to test for non-specific cytotoxic serum factors as described before [2]. Absorption of the sera with equivalent quantities of Q-fever antigen of phase I and II were carried out and the sera were tested accordingly,

56 B. Maisch: Rickettsial perimyocarditis

Q-fever myocarditis

onset of symptoms

o ÷ ÷ ( * ) - - -

320 ÷ + (+) . ~ _ _

- - 4 - + ÷ - - - -

121

"~ 80 ~0

4j 2 6. 7'7 g.7. " lg.sB0 islu82

U.W. 37y.

[ Clinical data I

Cephalgia

Pteuropneumonia

Tachyarrhythmia Atrial flutter

• O -fever CFT

Fig. 1. Clinical data during the acute phase and follow-up of the patient (UW, 37 years old, male) with Q-fever myocarditis. A fourfold increase in the complement fixation titer of phase II antigen occurred in the course of the disease

Assays of cytotxicity

Lymphocytotoxicity was assessed accordingly in the same assay system using isolated cardiocytes and the patient's lymphocytes in target cell/lymphocyte ratios of 1 : 10 and 1 : 50 as has been described [2]. The index of lymphocytotoxicity calculated in a similar manner ICILc = (ts0i):(ts0c) ] was rated abnormal if below 0.76 (mean - 2SD; normals 1.06 + 0.14; n = 32).

Assays of factors blocking or enhancing lymphocytotoxicity were carried out in parallel in a formal set-up to test antibody- dependent cellular cytotoxicity (ADCC), in which the patient's and control lymphocytes were incubated with the patient's serum. De- tails of the assay have been described [2].

Statistics

Results are given as mean (x) + 1 SD.

Results

Case report

A 37-year old patient with acute perimyocarditis due to a Q-fever infection (phase I antibody negative, phase II antibody of the IgM and IgG types with increasing positive titers) was followed up for 2½ years.

Clinical course

Two weeks prior to admission to hospital, the patient had experienced fatigue, sweats, cephalgia, fever of up to 40°C, and coughing. A first X-ray revealed infiltration of the sixth segment of the lower left lobe. Body temperature normalized under treatment with Amoxicillin (3 x 750 mg); the infiltration dis- appeared, whereas coughing, sweats, and fatigue per- sisted (Fig. 1).

When admitted to hospital, sudden precordial pain, tachyarrhythmia, and cephalgia were present. The heart rate ranged between 128 and 144 beats/ min.

Table 1. Clinical and Serological data

Admittance Follow-up

Sedimentation rate 65/104 7/23 Hemoglobin 11.3 g/dl 14.2 g/dl Leukocytes 15 700/mm 3 5 600/mm 3 Phase I antibodies (Rickettsia Negative Negative

burnetii) Phase II antibodies (IgG type) 1 : 320 1 : 160 Phase II antibodies (IgM type) Positive Low/negative

X-rays. Infiltration located on the dorsotateral right lobe during the first 2 weeks disappeared gradually leaving a pleural adhesion. Two lower lobes showed pleural effusions. ECG. Atrial flutter (300/min), a ventricular rate of 124/min, and nonspecific ST-segment depression were revealed in ECGs. Echocardiography. Pericardial thickening (11 mm) but no valvular heart disease or endocardial vegeta- tions were detected by echocardiography. There was fractional shortening of 18% in the beginning when septal movement was abnormal and 42% at the end of treatment when septal motion had returned to normal. Left ventricular end-diastolic dimensions diminished from 51 to 45 mm in the follow-up. The relevant laboratory findings are listed in Table 1.

After treatment with tetracyclin, the patient im- proved, but atrial fibrillation and pericardial fibrosis persisted.

Immunological study

Antimyolemmal antibodies (AMLA) against intact human and rat cardiocytes of the IgM and IgG types were seen initially in titers of up to 1 : 320 (Figs. 2, 3) and decreased in titer gradually during follow-up. Cytolytic serum activity (antibody-mediated cytoly- sis) was initially high (index 0.52) and normalized

R Maisch: Rickettsial perimyocarditis

during the following 2½ years (Fig. 3). No cytolysis was present when complement was omitted in the assay.

Only marginal lymphocytotoxicity and antibody- dependent cellular cytotoxicity were demonstrated in the acute stage with an index of 0.73 and on one occasion 2½ years later (Fig. 4). In the follow-up, phase 2 antibody, AMLA, and antibody-mediated cytolysis decreased, although an anamnestic titer of

57

A M L A and residual antibody-mediated cytolysis re- mained (Fig. 1).

Absorption experiments

By absorption of the patients sera with phase I and II Q-fever antigens on the one hand and cardiocytes and control antigens on the other, it could be demon- strated that ant imyolemmal fluorescence and cytoly- tic serum activity were completely absorbed out with phase II antigen and isolated heart cells only, thus indicating cross reactivity between the phase I1 anti- gen of Rickettsia burnetii and the myolemma of iso- lated heart cells (Table 2).

Discussion

Fig. 2. Demonstration of antimyolemmal antibodies (together with antibodies against the Z bands) with isolated rat and human cardio- cytes. Serum dilution 1 : 20, FITC-labeled antihuman trivalent anti- serum. F(ab)2-fragments from the goat, Medac

Cardiac involvement in Q-fever is uncommon but has been described either as myocardial or endocardial disease [8, 10, 16, 17, 18]. Our patient, living in an area endemic with Q-fever infections, was the only one with overt cardiac manifestations. The symptoms in- dicated three stages of the disease: In the initial period [i], nonspecific symptoms and pneumonia (pre- hospital phase) were predominant ; in a second stage, cardiac symptoms (atrial flutter, tachyarrhythmia, and conduction disturbances) prevailed, which can either be interpreted as relapse [10] or as a second

Antimyolemmal antibodies and antibody mediated cytolysis

5J- ~ , / ",

?o 216. 7.?

~' [ Immunot0gicat I LXA | data t

controls Uw Q-fever Antimyolemmat

A /,, ~ O antibod~slAMLA)

Antibody mediated ,it ,,N cytotysis

~ ,,,,, n o ~ , ~ l i l ; o d i ¢ ~

18~0o. 15.17.B2 O-fever controls n=10

Fig. 3. Titer of antimyolemma[ anti- bodies (AMLA) and corresponding cytolytic serum activity as assessed by the index of cardiocytolysis (CIAMc), There was significant cytolysis present in the acute stages I and 11, but none or very little in the follow-up (III). No cardiocytolysis could be detected in Q-fever controls without AMLA (open circles or open trianqles) when compared with the patient shaded triangles for cytolytic serum activity or antibody-mediated cytolysis)

Lymphocytotoxie reactions against cardiocytes

1,0

5 x O

o,8 "6

.L~

T z 6. ................ 7'7. " ,8 8.8o ,s.fi.sz

U.W. c~trols In= 50) • 0 LC

A /', LCBF (ADCC)

Fig. 4. Only marginal lymphocytoto- xicity (LC) and borderline antibody- dependent cellular cytotoxicity (ADCC) and no lymphocytotoxicity blocking factors (LCBF) could be detected initially and in the follow-up of the patient

58 B. Maisch: Rickettsial perimyocarditis

Table 2. Effect of absortion with cardiocytes and Q-fever antigens (phase I and II) on the titer of AMLA and on cytolysis serum activity

Means of

absorption

AMLA titer Antibody-mediated cytolysis (index)

Before After Before After absorption absorption absorption absorption

Cardiocytes 1 : 320 0 0,52 0.96 Phase II antigen 1 : 320 0 0.52 0.82 Phase I antigen 1 : 320 1 : 40 0.52 0.59 Control antigen 1 : 320 1 : 320 0.52 0.54

disease due to secondary immunopathogenesis, as has been demonstrated for rheumatic fever [11] and some forms of viral perimyocarditis [2, 14]. After this period, there was a gray-zone transition to the chronic or healing phase III with some residual symp- toms but little, if any, inflammatory activity.

The immunological data, particularly the humoral effector mechanisms ofcytolytic antimyolemmal anti- bodies, are well compatible with a secondary immu- nopathogenesis, since they were still present in the chronic stages II/III. Secondary immunopathogenesis is indicated by cytolytic, complement-fixing, antimyo- lemmal antibodies in vitro. The degree of cytolysis correlated with the AMLA titers. Cytolysis could be abolished by absorption of sera with Coxiella burnetii phase II antigen only, indicating that this antibody is cross reactive with the sarcolemma and is comple- ment fixing and cytolytic. Phase i antigen, to which no antibody response occurred in our patient, and control antigen did not abolish the cytolytic reactiv- ity. Since target cells different from cardiocytes, e.g., hepatocytes, were not lysed by the patient's serum, the antibody response is likely to be cardioselective. It can, therefore, be concluded that cross-reactive, cyto- lytic, cardiospecific, antimyolemmal antibodies are operative in the chronic phase of Q-fever carditis, as has been demonstrated in Coxsackie B, influenza, and mumps myocarditis [2].

Only initially and during the follow-up period could minor direct lymphocytotoxicity be demon- strated, indicating target cell-specific lysis, which does not underly the restriction of lymphocyte identifica- tion by major histocompatibility complex antigens as required for cytotoxic T cells. Either enhanced natural killer cell activity [19], which seems unlikely from studies with NK-cell activity in acute carditis [20], or increased cardiocytotoxicity by broad but still target-selective effector mechanisms, perhaps due to a subpopulation of NK cells or monocytes, may be re- sponsible [21].

At present, the role of cytotoxic T lymphocytes in Q-fever myocarditis cannot be further elucidated since autologous vital cardiac target cells are not yet available. From animal experiments [22, 23], a cyto- toxic T-cell response which is restricted by simulta- neous identification of the major histocompatibility complex has to be postulated as well.

References t. Maisch B, Berg PA, Kochsiek K (t980) Autoantibodies and

serum inhibition factors in patients with myocarditis. Klin Wo- chenschr 58:219-225

2. Maisch B, Trostel-Soeder R, Stechemesser E, Berg PA, Kochsiek K (1982) Diagnostic relevance of humoral and cell mediated immune reactions in patients with acute viral myocarditis. Clin Exp Immunol 48" 533-545

3. Maisch B, Maisch S, Kochsiek K (1982) Immune reactions in tuberculous and chronic constrictive pericarditis--Clinical data and diagnostic significance of antimyocardial antibodies. Am J Cardiol 50:1007-1013

4. Berg PA, Brand H, Marker A (t973) Nachweis und Bedeutung von Antiktrpern gegen homologes Gefhgendothel bei Leber- und anderen Krankheiten. Verh Dtsch Ges Inn Med 9:649 652

5. Martin M, Betzon A (1960) Curable acute primary carditis due to rickettsia prowazekii. Presse Med 68:1253-1254

6. Murphy AM, Field PF (1970) The persistance of complement fixing antibodies in Q-fever (Coxiella burnetii) after infection. Med J Aust 1148-1150

7. Thong YH, Hensen SA, Vincent MM, Fuccillo DA, Stiles WA, Bellanti JA (1976) Use of cryopreserved virus-infected target cells in a lymphocytotoxicity 51 Cr release microassay for cell- mediated immunity to cytomegalovirus. Infect Immun 13: 643- 657

8. Butterworth S (1970) Electron microscopic findings in Q-fever. J Clin Pathol 23:377

9. Kazar J, Schramek S, Brezinar R (1977) Analysis of serum immunoglobulins in a patient with chronic Q-fever and endo- carditis. Bratisl Lek Listy 67: 109- 114

I0. Mitchell R, Grist NR, Bazar G, Kenmuir ACF (1966) Path- ological, rickettsiological and immunofluorescent studies of a case of Q-fever endocarditis. J Pathol Bact 191:317 322

11. Kaplan MH, Meyeserian M (1962) An immunological cross- reaction between group a streptococcal cells and human heart tissue. Lancet I: 706-710

12. Coons AH, Kaplan MM (1950) Localization of antigens in tissue cells. I. Improvements in a method tbr the detection of antigen by means of fluorescent antibody. J Exp Med 91:1-13

13. Powell R, Twist VW (1976) Rapid technique for the isolation and purification of adult cardiac muscle cells having respiratory control and a tolerance to calcium. Biochem Biophys Res Com- mun 72:327-333

14. Maisch B, Trostel-Soeder R, Berg PA (1981) Assessment of antibody mediated cytolysis of vital adult cardiocytes isolated by centrifugation in a continuous gradient of Percoll TM in patients with acute myocarditis. J Immunol Methods 44:159 169

15. Maisch B (1985) Surface antigens of the adult heart cells and their use in diagnosis. Basic Res Cardiol 80 (Suppl 1): 47 52

16. Barraclough D, Popert AJ (1975) Q-fever presenting with paroxysmal ventricular tachycardia. Br Med J 2:423 424

17. Woodward TE, McCrumb PR, Carey TN, Togo Y (1960) Viral and rickettsial causes of cardiac disease, including the Cox- sackie virus etiology of pericarditis and myocarditis Ann Intern Med 53:1130-1150

B. Maisch: Rickettsial perimyocarditis

18. Constantinidis K, Jenkins JP (1979) Chronic Q-fever endocar- ditis. Practitioner 22: 533-1150

19. Pearlmann P, Perlman H (1971) Cytotoxic lymphocytes. Mech- anisms of activation and target-cell destruction. Int Arch A1- lergy 41: 36t --39

20. Maisch B (1985) Cellular effector mechanisms in acute carditis and postmyocarditic dilated heart disease. Basic Res Cardiol (in press)

21. Biilowius U, Maisch B, Klopf D, Schmier K, Hiby A, Schunk

59

D, Kochsiek K (1983) Lymphozytensubpopulationen bei akuter Perimyokarditis, sekund/iren und prim/itch dilatativen Her- zumskelerkrankungen. Z Kardiol 72 (suppl 11:16

22. Paque RE, Gauntt CJ, Nealon TJ, Trousdale MD (t978) As- sessment of cell-mediated hypersensitivity against Coxsackie virus B3 viral-induced myocarditis utilizing hypertonic salt ex- tracts of cardiac tissue. J lmmunol 120:1672-1678

23. Huber SA, Lodge PA (1984) Coxsackievirus B-3 myocarditis in Balb/c mice. Am J Pathol 116:21-29