key: cord-1032768-tsb6gul8 authors: Riski, Hellevi; Hovi, Tapani title: Coronavirus infections of man associated with diseases other than the common cold date: 2005-12-06 journal: J Med Virol DOI: 10.1002/jmv.1890060309 sha: 3a2faeb2a5dd6380694b5ec430cd00616d8a3f1e doc_id: 1032768 cord_uid: tsb6gul8 About 14,000 paired sera, from patients with various types of acute infectious diseases with suspected viral origin, were screened by complement fixation against a wide set of viral antigens, including coronavirus OC43. A significant change in OC43 antibodies was recorded in 33 cases and a constant high titre, defined as a titre occurring in the respective age group in less than 1 % of all sera examined, was found in 45 cases. On the basis of careful retrospective analysis of hospital case records it was concluded that in 28 cases with an increase of OC43 antibody litres, and in two with titre decrease, a disease could be associated with an acute coronavirus infection. In 16 cases the disease was dominated by respiratory symptoms. Eight of these patients, four children and four adults, had pneumonia. Three of the eight pneumonia patients had, however, another concomitant infection, too. Four patients had neurological symptoms, one had severe perimyocarditis, and in five cases fever was the only symptom recorded. Among the patients with a statistically significant high titre of OC43 antibodies, there were 14 cases where a suggestive association with a disease could be envisaged on the basis of hospital records. Five of these patients had pneumonia. These results suggest that human coronaviruses, so far considered only as one group of causative agents of the common cold, may also be associated with other and more severe diseases in all age groups. . Lower respiratory tract infections have rarely been described. Coronavirus strain 229E was recovered from nasal washes of two children with acute phase pneumonia [McIntosh et al, 19741 . During an outbreak of coronavirus infections among recruits, four cases with pneumonia showed seroconversion to OC43 virus antigen [Wenzel et al, 19741. Problems in diagnosing human coronavirus infections are well known, namely the failure of most strains to grow in regular cell cultures used for virus isolatioii as well as the technical difficulties in preparing sufficient amounts of antigen other than OC43 for routine serological screening. Hence it is probable that our understanding of the real clinical significance of human coronavirus infections is still limited and superficial. Indeed, electron microscopic studies have revealed coronavirus-like particles in patient tissues or secretions in a variety of nonrespiratory disorders of man, such as acute hepatitis [Holmes et al, 19701, chronic active hepatitis [Zuckerman et al, 19701, endemic nephritis [Apostolov et al, 19751 , acute nonbacterial gastroenteritis [Caul and Clarke, 197.51 , and lymphatic neoplasia [Arnold, 19781. It is not known, however, whether these associations really reflect a causal relationship between the coronaviruses and the indicated clinical diseases. In the routine diagnostic laboratory of this department more than 4,000 paired sera, from patients with various infectious diseases of suspected viral origin, are annually screened by the complement fixation test against a standard set of viral antigens. By using a wide set of antigens it is theoretically possible to discover new disease associations for viruses included in the set. We have had coronavirus OC43 antigen in this screening lest since March 1977 and report in this paper a review of the clinical cases so far associated with serologically documented OC43 infections. From March 1977 to March 1980 paired serum samples from about 14,000 patients were screened at the routine diagnostic laboratory of this department. The patients in question were living mainly in the southern part of Finland, and in a great majority of cases they had been referred to a hospital because of a moderate or severe illness with suspected viral origin. All sera were screened by the complement fixation (CF) test against the following agents: adenovirus, coronavirus OC43, coxsackie ims type B5, cytomegalovirus, herpes simplex virus type 1, influenza viruses A and B, measles, mumps, parainfluenza types 1 and 3, poliovirus type 2, respiratory syncytial virus, rotavirus, and varicella-zoster as well as chlamydia, mycoplasma pneumoniae, and toxoplasma. A crude OC43 virus preparation grown in mouse brain [Riski et al, 19771 was used as antigen in the CF antibody test carried out by a standard microwell method [Palmer and Casey, 1969; Hovi et al, 19791 . A titre of 1 :8 or greater was considered positive. All of the serum pairs with a significant change or a high level of OC43 CF antibodies were subsequently tested with a control antigen prepared from uninfected mouse brain, in order to detect possible false-positive reactions. Most of the serum pairs with an antibody increase in the CF test were also tested with OC43 haemagglutination inhibition (HI) [Hovi, 19781 and with the haemolysis-in-gel (HIG) test [Riski et al, 19771 . Results ob-tained with CF and HIG tests were in good correlation, whereas several cases showing a definite increase of CF antibodies had a constant low HI-titre. During the screening period all cases with significant (fourfold or greater) changes in OC43 CF antibody titre as well as those with a constant high titre were recorded. The limit for the high titre, 64 or 128, was determined separately for each age group by a computer analysis of the titre distribution of 8,448 serum samples tested at this department in 1978 [Ukkonen et al, in preparation] . Titres seen in less than 1% of all sera tested in the respective age group were considered high in this study (>128 or 264). The symptoms of 78 clinical cases thus collected were retrospectively carefully analysed according to case records obtained from the hospitals. Sera from patients with a constant high titre were included in the material only if the first sample had been taken later than 2 weeks after the beginning of the symptoms. Sera from high-titre patients with immunomodulating disease or treatment, such as chronic liver disease or kidney transplantation and steroid treatment, respectively, were excluded from the material. Neither were those high coronavirus titres included where evidence was obtained of another, more probable cause for the disease, eg, bacteremia in a case of pneumonia, nor those where high titres were measured against more than one other antigen. After this critical evaluation, sera from 14 cases with a constant high titre were accepted in the material. Because the temporal coincidence of the infection and the disease in these cases, as well as in cases showing titre decrease, cannot be proven, they are listed separately from those showing a titre increase. Patients showing an increase of coronavirus OC43 antibodies were distributed throughout the different age groups (Table I) . However, there was a remarkable accumulation of cases in the groups below 7 years of age. More than one-third (1 1/28) of the titre increases but only 12% of serum pairs studied came from patients in these age groups. High titres that we considered significant, however, showed a different age distribution with only one patient younger than 7 years (Table 11) . A great majority (20/28) of the patients with an increase of OC43 antibodies were male ( Table I) . Because of the low total frequency of documented coronavirus infections in the material and the way of collecting the material, this uneven sex ratio should not be taken as evidence for differential susceptibility to coronavirus infection-associated clinical disease. spiratory infections (16/28 and 12/16, respectively). About half of these were common colds and most of the other patients had pneumonia. In two pneumonia cases there was a concomitant increase of mycoplasma antibodies, and in one case there was a rise of influenza B antibodies. All the pneumonias associated with coronavirus infections were, on the clinical grounds, "atypical," ie of suspected viral origin, and all the patients recovered rapidly. Most of the pneumonia patients had been healthy before. Coronavirus infections were, however, also seen in association with exacerbations of chronic respiratory diseases such as bronchial asthma and chronic bronchitis (Tables I and 11 ) [see also Henigst, 19741 . 5. 6. 7. 8. 10. 11. 13. 14. 16. Six patients had neurological symptoms together with signs of an acute infection (Tables I and 11 ). Like the pneumonias, neurological symptoms disappeared rapidly resulting in complete recovery of the patients. Some of the cases are described in more detail in the following. Case 17. A 3-year-old boy had a typical serous meningitis with a moderate increase of CSF leukocyte counts (max 348 per mm3). Sixty-eight to ninety-four percent of the cells were lymphocytes. Cultures for bacteria from CSF specimens were negative. The neurological symptoms disappeared within 5 days under conventional care. This patient was found to have a concomitant increase in chlamydia CF antibodies, but did not receive antibiotics. The specificity of both corona and chlamydia antibodies was verified by tests using adequate control antigens. Case 19. A 17-year-old female was referred to a hospital because of high temperature and severe headache. Meningitis was excluded, however, by absence of neck stiffness and normal findings on examination of CSF specimens. The patient recovered within a week. Case 18. A 4-year-old boy who had been generally healthy before had not had acute infection-associated convulsions previously. After 2 days with typical symptoms of common cold he now had an attack of convulsions and unconsciousness and subsequent hemiparcsis for a few hours. Electroencephalography after the attack showed mild, nonfocal changes. The patient recovered fully in a few days. Case 21. This patient had a relatively severe perimyocarditis with typical signs and symptoms. Recovery was relatively slow with the patient still complaining of tiredness 3 months after the onset of the disease. These results indicate that coronavirus infections associated with diseases other than the common cold are not rare. On the basis of the available data it is, however, very difficult to judge if the documented infections really caused the associated diseases. Infection by another unidentified agent concomitantly with the serologically documented coronavirus cannot be fully excluded, which, of course, is always a problem in the search for specific diagnosis for a disease or symptom with several potential etiological agents. Our goal in the present study has been to record all possible symptoms and diseases coinciding with documented coronavirus infection whatever the cause-and-effect relation of the two phenomena is. With this goal in mind it is not possible to use any control materialfor instance, sera from healthy persons collected during the same time periodbecause subclinical coronavirus infections might have occurred in them as well. Concomitant virus isolations from the patients could have given further evidence for the possible cause-and-effect relation, but samples for virus isolation were taken from these patients only exceptionally and, furthermore, it would not be possible in practice to carry out attempts for coronavirus isolation in a material of this size. More straightforward approaches would be necessary to establish the potential role of human coronaviruses in the etiology of the observed diseases. The %ew" disease associations were, however, not unexpected. Coronavirus infection-associated pneumonia in man has been described before both in children [McIntosh et al, 19741 and in young adults [Wenzel et al, 19741 . Eight out of the 13 pneumonia cases (including both titre increases and high titres) in the present study were adults. Double infections were seen in four out of 13 pneumonia patients. Both mycoplasma and influenza B can cause pneumonia without concomitant coronavirus infection, and it remains unresolved what role, if any, the observed coronavirus infections had in the pathogenesis of these four pneumonia cases. Interesting were two cases with pleurodynia-like symptoms, so far almost exclusively associated with coxsackie virus infections, as well as a case with relatively severe perimyocarditis. fection and also comprised the other major group of symptoms in this study. No typical pattern of neurological symptoms could be ascribed to these cases. It seems possible, however, that a fraction of mild meningitis and radiculitis cases would be caused by coronaviruses. It is also worth mentioning that in seven of 44 cases prolonged fever was the only symptom noted in the patients. In four cases no specific symptoms were recorded for the period showing OC43 titre increase. These cases may represent subclinical infections or, more probably, possible symptoms of common cold were not recorded as the records were kept for other reasons, eg, for follow-up after a surgical operation. Neurological symptoms are associated at a certain rate with practically any viral in-In conclusion, these observations suggest that human coronaviruses, apart from causing a considerable part of common colds in man, may also be involved in the etiology of more severe diseases in all age groups. These results also prove the usefulness of the wide set of viral antigens in serological diagnosis of viral diseases, as compared to a strategy of using only "the most probable viruses" as antigens. It is unlikely that before these studies a coronavirus antigen would have been included in a "pneumonia virus collection" or, even less probably, in a set used in cases of suspected viral infections of the central nervous system. Evidence of a viral aetiology on endemic (Balkan) nephropathy Identic viral infections in four cases of malignant lymphoepithelioma Coronavirus propagated from patient with non-bacterial gastroenteritis Coronavirus infections in working adults: Eight year study with 229E and OC43 Occurrence of antibody against coronavirus (OC 43) in the healthy population and in patients with a disease of the respiratory tract and bacterial respiratory infections with exacerbations of wheezing in young asthmatic children Coronavirus infection in acute lower respiratory tract disease in infants Standardized Diagnostic Complement Fixation and Adaptation to Microtest Three-year study with coronavirus strains OC43 and 229E Viruses associated with acute respiratory infections 1961 -71 Nonspccific inhibitors of coronavirus OC43 haemagglutination in human sera Seroepidemioiogic survey of coronavirus (strain 229E) infections in a Coronaviruses: A comparative review. Current Topics in Microbiology and Immu The association of viral Antibodies to human coronavirus OC43 measured by Wenzcl RP Prcsencc of particles other than the Australia-SH anti