Enteroviruses Can Persist with or without Active Viral Replication in Cardiac Tissue of Patients with End-Stage Ischemic or Dilated Cardiomyopathy

Transcription

1 1222 CONCISE COMMUNICATION Enteroviruses Can Persist with or without Active Viral Replication in Cardiac Tissue of Patients with End-Stage Ischemic or Dilated Cardiomyopathy Laurent Andréoletti, 1,7,a Thomas Bourlet, 5 Donatien Moukassa, 2 Laurent Rey, 1 David Hot, 4 Yanwen Li, 5 Valérie Lambert, 1 Bernard Gosselin, 2 Jean-François Mosnier, 6 Czelas Stankowiak, 3 and Pierre Wattré 1 1 Laboratoire de Virologie, 2 Département d anatomopathologie, and 3 Centre de Chirurgie Cardiovasculaire, Hôpital cardiologique Centre Hospitalier Universitaire, and 4 Département de microbiologie des Ecosytèmes, Institut Pasteur de Lille, Lille; 5 Laboratoire de Virologie and 6 Département d anatomopathologie, Centre Hospitalier Universitaire, St. Etienne; and 7 Département de Virologie, Hôpital Broussais, Paris, France To investigate enterovirus replication versus persistence in end-stage cardiac diseases, endomyocardial biopsies from explanted hearts of 70 patients with idiopathic dilated cardiomyopathy (IDCM), 64 patients with chronic coronary disease (CCD), and 45 donors of healthy hearts (controls) were examined by reverse transcriptase polymerase chain reaction for genomic and antigenomic enterovirus RNA and by VP1 antigen immunohistochemistry. Enterovirus genome was detected in 25 of 70 patients with IDCM and in 21 of 64 patients with CCDs (35.7 vs. 32.8%, respectively; P p.12). Of the 46 patients positive for genomic RNA, only 3 exhibited antigenomic RNA and VP1 antigen that demonstrated active viral replication, whereas 43 had latent infection characterized by the absence of antigenomic RNA associated with or not with VP1 antigen expression. No viral component was detected in control subjects. The findings demonstrate that a small percentage of patients with end-stage chronic cardiac diseases had active enterovirus replication in their myocardium. Enteroviruses (EVs) of the Picornaviridae family, particularly coxsackie B viruses, possess cardiotropic activity and are important causes of myopericarditis in children and adults [1]. Although most EV illnesses are subclinical, acute myocardial infection can induce severe arrhythmias and sudden cardiac death and lead to the development of a chronic form of myocarditis, dilated cardiomyopathy (DCM), which is the second leading etiological cause of heart transplantation [1, 2]. Recent polymerase chain reaction (PCR) studies in human myocardium have provided evidence that EV infection was detectable in acute and chronic myocarditis, as well as in end-stage idiopathic dilated cardiomyopathy (IDCM) [1 4]. Furthermore, EV RNA was found in explanted hearts from 17% 39% of patients undergoing transplantation because of end-stage chronic coronary diseases (CCDs) [4, 5]. These findings, together with serological studies that demonstrate concomitant increases in an- tibodies to coxsackieviruses B at diagnosis of myocardial infarction, suggest the involvement of EVs in acute and chronic ischemic heart diseases [6]. Together, these data suggest that the development of IDCM and CCDs could be associated with the EV infection of endomyocardium. However, the precise role of EVs in the pathogenesis of these chronic heart diseases remains to be investigated. In the present study, the involvement of persistent EV infection in chronic cardiomyopathy was investigated by seminested reverse transcriptase (RT) PCR detection of genomic virus RNA in myocardium from patients with end-stage IDCM or CCD and from donors with healthy hearts (control group). Moreover, using a specific seminested RT-PCR detection of antigenomic viral RNA strands and a viral capsid protein VP1 immunostaining assay, we analyzed endomyocardial virus replication activity at the end stage of chronic cardiac diseases. Received 20 March 2000; revised 24 May 2000; electronically published 8 September Consent was obtained from patients or their relatives. Financial support: Centre Hospitalier Régional Universitaire et Conseil Regional Nord-Pas de Calais (no ). a Present affiliation: Laboratoire de Virologie, Hôpital Europeen Georges Pompidou, Université Pierre et Marie Curie (Paris VI), Paris, France. Reprints or correspondence: Dr. L. Andréoletti, Laboratoire de Virologie, Hôpital Européen Georges Pompidou, 20, rue Leblanc, Paris Cedex 15 The Journal of Infectious Diseases 2000;182: by the Infectious Diseases Society of America. All rights reserved /2000/ $02.00 Patients and Methods Transplanted patient groups. Seventy patients (61 male and 9 female; mean age, 42.5 years) with end-stage IDCM, according to the World Health Organization criteria [7], were studied. For each patient, a ventricular endomyocardial biopsy was obtained at the time of cardiac transplantation, quickly frozen in liquid nitrogen, and stored at 80 C. For each patient, a part of the myocardial biopsy was histopathologically examined, using hematoxylin-eosin stain. Histologic findings have shown a focal fibrosis and the absence of inflammatory cells, which are DCM-like features [2].

2 JID 2000;182 (October) Enteroviruses and Chronic Myocardial Diseases 1223 Figure 1. Representative detection of enteroviral genomic and antigenomic RNA by reverse transcriptase polymerase chain reaction (RT- PCR) assays in endomyocardial tissue of patients with end-stage chronic heart diseases. A1 and A2, Sensitivity of genomic enteroviral and antigenomic enteroviral RNA detection, respectively, by seminested RT-PCR. Products were visualized by agarose gel electrophoresis. Lanes 3 8, Products derived from the amplification of RNA extracted from 10-fold dilution positive or negative RNA strands transcribed from a complete CVB3 cdna sequence in 100 mg of murine heart homogenate, corresponding to (lane 2) and 2 (lane 8) CVB3 RNA copies [8]. Seminested 362-bp RT-PCR product was detected in as few as 20 copies of positive (A1, lane 7) or negative (A2, lane 7) CVB3 RNA strands. Lane 1, Seminested RT-PCR reagent blank (sterile water). A3 and A4, Plus-strand (genomic) and minus-strand (antigenomic) enteroviral RNA detection, respectively. Amplification was done in 100 mg of heart tissue homogenate from adult patients with idiopathic dilated cardiomyopathy (IDCM; lanes 4 8), adult patients with chronic coronary disease (CCD; lanes 9 11), and donors with healthy hearts (lanes 12 14). A 362-bp enteroviralspecific fragment can be seen in lanes 4, 6, and 7 9 but not in lanes 5 and Lane 1, Positive control obtained by amplification of positive (A3) or negative (A4) RNA strands that were transcribed from a complete CVB3 cdna sequence; high no. ( copies) of plus (A3) or minus (A4) strands induced the co-amplification of the first RT-PCR product (435 bp) during the second PCR run (seminested RT-PCR assay). Lane 2, Semi-nested RT-PCR reagent blank. Lane 3, Strand-specific enteroviral RNA detection; A1, negative signal with negative CVB3 RNA strands; A2, negative signal with positive CVB3 RNA strands. A5, Glyceraldehyde phosphate dehydrogenase (G3PDH) RT-PCR products (983 bp) were visualized by agarose gel electrophoresis. Lane 1, PCR amplification of complete cdna G3PDH sequences (provided by Clontech, Palo Alto, CA) and was used as the positive PCR control. Lane 2, RT-PCR reagent blank. Lane 3, Positive G3PDH RT-PCR detection in a heart sample from a healthy heart donor. Lanes 4 14, Positive G3PDH mrna detection in heart tissue of IDCM patients (lanes 4 8), CCD patients (lanes 9 11), and healthy heart donors (lanes 12 14). For these patients (lanes 4 14), results of seminested RT-PCR detection of enteroviral genomic and antigenomic RNA are shown in A3 and A4. CCD. Sixty-four patients (57 males and 7 females; mean age: 62 years) with ischemic cardiac disease, defined as significant coronary heart disease requiring cardiac transplantation, were studied. Endomyocardial biopsies were obtained according to the protocol of the IDCM patient group. For each patient, part of the endomyocardial biopsy was histopathologically examined to confirm the clinical finding of ischemic heart disease [4]. Control group. The control group consisted of myocardial biopsies from 45 young adult organ donors who died suddenly as a result of an accident or suicide. None of these subjects were known to have cardiac pathology. All control samples were from the ventricle and had no histologic abnormalities. Cell culture. A part of each endomyocardial biopsy was homogenized in 1 ml of MEM. A 150-mL aliquot of these crude extracts was inoculated in duplicate to Hep2 and MRC-5 cells for 14 days with 1 subculture passage [8, 9]. RNA extraction and RT-PCR procedure. Native RNA was extracted from a 100-mg myocardial biopsy specimen with RNAzole (Bioprobe Systems, Montreuil Sous Bois, France), according to the manufacturer s instructions [8]. An EV seminested RT-PCR assay was done as described elsewhere [9]. To obtain a broad specificity for the genus Enterovirus, 3 primers described for annealing in the highly conserved sequences within the 5 noncoding region (5 NC) were used [9, 10]: P2 (nucleotide position of CVB3-Nancy genome: ), 5 -CAAGCA- CTTCTGTTTCCCCGG-3 ; P3 (nucleotide position of CVB3- Nancy genome: ), 5 -ATTGTCACCATAAGCAGCCA-3 ; and EV1 (nucleotide position of CVB3-Nancy genome: ), 5 -CTTGCGCGTTACGAC-3 ). cdna for plus- or minus-strand RNA was specifically synthesized by use of sense (P2) or antisense (P3) primer (annealing temperature, 45 C) [9]. The cdna was heated to 95 C for 30 min to inactivate the RT activity and, therefore, to avoid a nonspecific RT activity during the following seminested RT-PCR procedures, as described elsewhere [11]. Classical PCR amplification of cdna was done, using primers P2 and P3, as described elsewhere by Leparc et al. [9]. A second PCR run was done, using primers P2 and EV1 [9]. Each heart sample was PCR tested in duplicate. Control positive or negative strands of RNA were transcribed from a complete CVB3 cdna sequence (CVB3-M1), which was

3 1224 Andréoletti et al. JID 2000;182 (October) Table 1. Results of detection of genomic enterovirus RNA, antigenomic enterovirus RNA, and enterovirus VP1 capsid antigen protein in the endomyocardial tissue of patients with idiopathic dilated cardiomyopathy (IDCM) or chronic coronary disease (CCD) requiring heart transplantation and in healthy heart donors without known cardiac disease. Patients No. tested Enteroviral markers positive in endomyocardial biopsies, no. (%) of patients positive Plus-strand RNA, minus-strand RNA, VP1 capsid protein Plus-strand RNA, VP1 capsid protein Plus-strand RNA No markers IDCM 70 2 (2.9) 15 (21.4) 8 (11.4) 45 (64.3) CCD 64 1 (1.6) 8 (12.5) 12 (18.7) 43 (67.2) Donors (100) NOTE. Genomic enterovirus (plus strand) and antigenomic (negative strand) RNA were detected by seminested reverse transcriptase polymerase chain reaction. Enterovirus VP1 capsid protein was detected by endomycardial immunostaining. inserted in a pbluescript (SK ) vector (Stratagene, La Jolla, CA) [8]. For all myocardial specimens, the glyceraldehyde phosphate dehydrogenase (G3PDH) mrna was amplified in an RT-PCR reaction [4] and used to verify the quality of RNA extraction. A 15-mL aliquot of each amplified RT-PCR product was subjected to classic agarose gel electrophoresis (figure 1) [4]. Immunohistochemistry. Mouse monoclonal anti EV antibody solution directed against capsid protein VP1 was used as the primary antibody (5D8/1 monoclonal antibody; DAKO, Copenhagen) at a 1:500 dilution. An immunoperoxidase revelation was then performed (Envision/HRP; DAKO) according to published protocols [12]. As a control, the primary antibody was replaced with diluent only. In addition, EV-positive or -negative murine myocardial sections from mice with CVB3-induced chronic myocarditis were processed in a manner similar to that for control subjects, as described elsewhere [8]. Statistical analysis. A x 2 test was done using Epi Info 6 (Centers for Disease Control and Prevention, Atlanta, and World Health Organization) statistical analysis software. Results of statistical analysis were considered significant at P!.05. Results The EV seminested RT-PCR procedure was done using a set of primers that allowed for the amplification of a specific fragment from 60 of 66 human EV serotypes [10]. The sensitivity of the RT-PCR assay was assessed by limit detection of the signal in serial 10-fold dilutions of positive or negative RNA strands transcribed from a complete CVB3 cdna sequence that was added to homogenized murine heart tissue. Agarose gel electrophoresis detected 20 copies of genomic or antigenomic RNA per 100 mg of cardiac tissue (figure 1A1 and 1A2). For each human endomyocardial biopsy tested ( n p 179), the G3PDH mrna was successfully amplified and used as a control to verify removal of PCR inhibitors from RNA extracts (figure 1A5). Using an endomyocardial seminested RT-PCR assay, we detected EV genome in 25 (37.5%) of 70 patients with IDCM and 21 (32.8%) of 64 patients with CCD; none was detected in 45 healthy heart donors (table 1; figure 1A3). No significant difference was observed in the genomic RNA detection rate between IDCM and CCD patients (37.5 vs. 32.8%; P p.12, x 2 test). To assess whether infectious EV particles could persist in the cardiac tissue at the end stage of cardiac diseases, we homogenized cardiac tissue and inoculated it into Hep2 and MRC-5 cell cultures. None of the 134 patient biopsies or 45 control hearts samples was culture positive 14 days after inoculation, which thus demonstrates the absence of cultivable EV particles. To determine whether endomyocardial EV persistence is associated with viral replication activity, we examined heart biopsies by use of a specific seminested RT-PCR detection of antigenomic virus RNA strands and by viral capsid protein VP1 immunostaining assay. Neither antigenomic RNA nor viral protein was detected in the samples from the control group consisting of adult heart donors (table 1). Of the 46 patients positive for genomic RNA, only 3 (6.5%; 2 with IDCM and 1 with CCD) were positive for antigenomic RNA, demonstrating a genomic viral replication (figure 1A4), whereas 43 (93.5%) of them demonstrated a latent viral infection characterized by the absence of antigenomic RNA (table 1). Myocytes from 17 IDCM and 9 CCD patients immunostained positive for VP1 (table 1). Detailed histopathologic examinations are shown in figure 2. Of interest, the samples from the 3 patients positive for both genomic and antigenomic RNA by PCR also exhibited the expression of VP1 antigen, which demonstrates active endomyocardial viral replication (table 1). Of the 43 patients with a latent viral infection (presence of genomic viral RNA without antigenomic viral RNA), 15 IDCM and 8 CCD case patients were positive for VP1 antigen, which suggests that a viral protein synthesis activity occurred during persistent endomyocardial viral infection and perhaps generated limited numbers of infectious or defective virus particles (table 1). Discussion In this study, the controversial association between persistent EV infection and chronic cardiac diseases was investigated by molecular and immunohistochemical detection of virus compo-

4 JID 2000;182 (October) Enteroviruses and Chronic Myocardial Diseases 1225 Figure 2. Detection of enterovirus capsid protein VP1 in ventricular endomyocardial tissue sections by immunohistochemistry. A, Positive immunohistologic detection of VP1 viral capsid protein in ventricular tissue sections from idiopathic dilated cardiomyopathy (IDCM) patients; only scattered myocytes (arrows) adjacent to chronic inflammatory lesions were positive for VP1 protein. B, Immunostaining of VP1 in heart tissue of a patient with IDCM. Viral antigens (yellow-brown) are localized in the cytoplasm of myocytes. C, Negative VP1 immunohistochemistry detection in heart tissue of an IDCM patient. D, Positive immunohistologic detection of VP1 viral capsid protein (yellow-brown) in ventricular tissue sections from chronic coronary disease (CCD) patients. Clusters of myocytes were positive for VP1 protein detection. Most positive myocytes were not adjacent to chronic histological tissue lesions. E, Intense immunostaining of VP1 in heart tissue of a patient with CCD. Signals (yellowbrown) localized in the cytoplasm of myocytes. F, Negative VP1 immunohistochemistry detection in heart tissue of a CCD patient. G, Immunohistological detection of VP1 viral capsid protein into ventricular tissue sections from NMRI nude (nu/nu) mice 31 days after CVB3 (Nancy) inoculation [8]. H, Negative immunohistologic detection of VP1 viral capsid protein into ventricular tissue sections from mock-infected NMRI nude mice [8]. All sections were counterstained with hematoxylin. Magnification: A, C, F, and H, 200; B and E, 400; D and G, 25. results were higher than those (12%) observed in similar IDCM patients by Keeling et al. [14]. However, our results disagree with 2 recent studies that failed to detect EV genomic RNA in heart samples of 38 and 22 DCM patients, respectively [15, 16]. Discrepancies between these data and our findings could be explained by differences in clinical and histological criteria for inclusion of DCM patients and by differences in geographic origin of the studied DCM patient groups. Moreover, we used an im- nents in myocardium from patients with end-stage IDCM or CCD and adult donors of healthy hearts (control group) [1 5]. Using a seminested RT-PCR [9], we detected EV genomic RNA in myocardial tissue of 35.7% of patients with unexplained (idiopathic) end-stage DCM but did not detect it in cardiac samples from healthy heart donors (table 1). Positive RT-PCR results were in agreement with those reported by Weiss et al. [13] and Pauschinger et al. [11] (45% and 36%, respectively), but positive

5 1226 Andréoletti et al. JID 2000;182 (October) proved EV seminested PCR assay with a sensitivity that was considered sufficient for studying clinical material, even from patients with chronic EV infections (figures 1A1 and 1A2) [9]. As previously described, the detection of the EV genome in association with a negative culture assay results in heart tissue sample suggests a persistent cardiac infection that could be implicated in the pathogenesis of IDCM. However, our failure to detect EV genomic sequences in 64.3% of patients with IDCM suggests that EV infection is not the first etiological cause of this chronic heart disease. Nevertheless, as suggested by Colston et al. [17], it is possible that, during the chronic physiopathologic process of IDCM, viral clearance by apoptotic destruction of infected myocytes occurs and results in negative EV RNA detection at the end stage of the disease. The percentage of patients showing EV RNA was similar in the CCD and IDCM groups, which points to a link between CCD and EV heart infection. In previous studies, 17% 39% of patients with CCD had EV RNA [4 5]. We do not yet know the significance of EV infection in CCD patients. EVs can infect in vitro human vascular endothelial cells, and myocarditis can be associated with microvascular spasm, small vessel obstruction, and myocardial reperfusion at the onset of infection [18]. EVs may be responsible for endothelial lesions that enhance vascularitis, atheromatosis, ischemia, and myocardial infarction [4, 6]. These findings, together with the detection of EV RNA in myocardium from patients with CCD in our study, point toward the involvement of EVs in acute or chronic ischemic heart disease. Using a specific RT-PCR detection of antigenomic virus RNA and a viral capsid protein VP1 immunostaining assay, we analyzed the endomyocardial virus replication activity in endomyocardial tissue from patients with end-stage IDCM or CCD. In our study, results for the antigenomic virus RNA and VP1 detection were positive in 2 IDCM patients and 1 CCD patient with positive results for the genomic viral RNA-PCR assay, which demonstrate that a small percentage of patients with end-stage CCDs had active EV replication in their myocardium. Moreover, 15 of 25 IDCM and 8 of 21 CCD case patients who were positive by viral genomic RNA PCR demonstrated a latent viral infection characterized by a viral genomic persistence either with or without VP1 capsid antigen expression (table 1). An earlier study reported heart samples from adult patients with myocarditis and from 8 adult DCM case patients positive for VP1 EV antigen [12]. To our knowledge, the current study is the second report on detection of endomyocardial viral antigen in end-stage IDCM and the first on the endomyocardial evidence of VPI capsid antigen in CCD cases who required cardiac transplantation. Our results further show a spatial correlation between VP1 immmunostaining and the chronic myocardial lesions in IDCM cases but not in cases of CCD (figure 2). These immunohistological findings suggest that the presence of a viral capsid protein synthesis activity may enhance the severity of cardiac lesions in the chronic physiopathologic process of IDCM but not in CCD. However, the implication of EVs as an enhancing factor in the pathogenic progress of CCD might be considered [4, 18]. Although we worked on large sections from explanted heart tissues, our immunohistological investigations did not allow for us to distinguish VP1-positive IDCM or CCD patients with or without EV replicative intermediates [11]. However, the detection of negative viral RNA allows for a molecular genetic differentiation between active viral replication and latent viral persistence, which should be of interest in future prospective clinical studies on the clinical course of EV chronic cardiomyopathies. In conclusion, this study provides further evidence to demonstrate that EVs can replicate in myocytes from IDCM and CCD patients to produce both capsid protein and RNA and, perhaps, to generate limited numbers of infectious or defective virus particles. Moreover, our findings demonstrate that EVs can persist in myocardium with or without genomic replication and viral protein synthesis activities, which could be involved in the pathogenesis of ischemic and dilated chronic heart diseases. Acknowledgments We are indebted to the clinicians from the Cardiologic Hospital of Lille and from the Centre Hospitalier Universitaire of St. Etienne for their active collaboration in the inclusion of heart-transplanted patients in this study. References 1. Baboonian CM, Davies J, Booth JC, Mc Kenna WJ. Coxsackie B viruses and human heart disease. In: Tracy S, Chapman NM, Mahy BWJ, eds. The coxsackie B viruses. Berlin-Heidelberg: Springer-Verlag, 1997: Archard LC, Bowles NE, Cunningham L. Enteroviruses RNA sequences in hearts with dilated cardiomyopathy: a pathogenic link between virus infection and dilated cardiomyopathy. In: Baroldi G, Camerini F, Goodwin JF, eds. Advances in cardiomyopathies. Berlin-Heidelberg: Springer-Verlag, 1993: Muir P, Archard LC. There is evidence for persistent enterovirus infections in chronic medical conditions in humans. Rev Med Virol 1994;4: Andréoletti L, Hober D, Decoene C, et al. Detection by PCR of enteroviral RNA in endomyocardial tissue of patients with chronic cardiac diseases. J Med Virol 1996;48: Petitjean J, Kopecka H, Freymuth F, et al. Detection of enteroviruses in endomyocardial biopsy by molecular approach. J Med Virol 1992;37: Roivainen M, Alfthan G, Jousilahti P, Kimpimaki M, Hovi T, Tuomilehto J. Enterovirus infections as a possible risk factor for myocardial infarction. Circulation 1998;98: Bradenberg RO, Chazov E, Cherian G. Report of the WHO/ISFC task force on definition and classification of the cardiomyopathies. Circulation 1980;64: Andréoletti L, Hober D, Becquart P, et al. Experimental CVB3-induced chronic myocarditis in two murine strains: evidence of interrelationships between virus replication and myocardial damage in persistent cardiac infection. J Med Virol 1997;52: Leparc I, Aymard M, Fuchs F. Acute, chronic and persistent enterovirus and poliovirus infections: detection of viral genome by seminested PCR amplification in culture-negative samples. Mol Cell Probes 1994;8: Zoll GJ, Melchers WJ, Kopecka H, Jambroes G, van der Poel HJ, Galama JM. General primer-mediated polymerase chain reaction for detection of

6 JID 2000;182 (October) Enteroviruses and Chronic Myocardial Diseases 1227 enteroviruses: application for diagnostic routine and persistent infections. J Clin Microbiol 1992;30: Pauschinger M, Doerner A, Kuehl U, et al. Enteroviral RNA replication in the myocardium of patients with left ventricular dysfunction and clinically suspected myocarditis. Circulation 1999;99: Li Y, Bourlet T, Andréoletti L, et al. Enteroviral capsid protein VP1 is present in myocardial tissues from some patients with myocarditis or dilated cardiomyopathy. Circulation 2000;101: Weiss LM, Movahed LA, Billingham ME, Cleary ML. Detection of coxsackievirus B3 RNA in myocardial tissue by the polymerase chain reaction. Am J Pathol 1991;138: Keeling PJ, Jeffery S, Caforio ALP, et al. Similar prevalence of enteroviral genome within the myocardium from patients with idiopathic dilated cardiopathy and controls by the polymerase chain reaction. Br Heart J 1992; 68: de Leeuw N, Melchers WJ, Balk AH, de Jonge N, Galama JM. No evidence for persistent enterovirus infection in patients with end-stage idiopathic dilated cardiomyopathy. J Infect Dis 1998;178: Crespo-Leiro MG, Hermida-Prieto M, Pena F, et al. Absence of enteroviral RNA in hearts explanted from patients with dilated cardiomyopathy. J Heart Lung Transplant 2000;19: Colston JT, Chandrasekar B, Freeman GL. Expression of apoptosis-related proteins in experimental coxsackievirus myocarditis. Cardiovasc Res 1998; 38: Conaldi PJ, Serra C, Mossa VF, et al. Persistent infection of human vascular endothelial cells by group B coxsackieviruses. J Infect Dis 1997;175:693 6.