Prospective serological and clinical studies on infants born in Kuwait with an elevated IgM in cord blood*

Bulletin of the World Health Organization, 63(4): 745 -750(1985) Cc) World Health Organization 1985 Prospective serological and clinical studies on infants born in Kuwait with an elevated IgM in cord blood* W. AL-NAKIB,1 L. V. DEVERAJAN, H. THORBURN,3 M. E. A. IBRAHIM,4 M. A. A. MOUSSA,5 H. HATHOUT,6 A. M. YOUSOF,7 & S. SOUFI8 Little is known about congenital viral and toxoplasmal infections in the developing countries. This study attempts to provide preliminary data on such infections among Arab mothers and infants in Kuwait. Babies born with an elevated IgM in cord blood were selected at birth and then studied prospectively during the first year of lifefor serological and clinical evidence of viral and toxoplasmal infections. There appeared to be a significant association between the elevation of total cord blood IgM and the selection of cases with cytomegalovirus infection. Demonstration of specific IgM in cord sera was attempted, whenever possible, to providefurther evidence ofcongenital infection and to exclude early postnatal infections. In 18.5% of cases there was an association between serological evidence of infection and certain clinical abnormalities. Apart from cytomegalovirus and rubella virus, no evidence was found for congenital infection by other viruses or toxoplasma in this study. Seroepidemiological studies conducted in Kuwait have shown that over 95% of 15-19-year old females had been exposed to rubella virus (RV), cyto- megalovirus (CMV), and alpha herpes virus 1 or 2 (HSV) infections at an earlier age and hence were immune by the time they were child-bearing (1); 58% of the same women had also acquired a toxoplasmal infection. These data suggest that primary congenital infection, especially by RV or CMV, is unlikely to play a significant role in the development of malform- ations in this population; and still the incidence of congenital malformations during the first week of life was comparable to that in other parts of the world, as reported in a previous study (2); moreover, no genetic or ethnic factors could be found to account for these malformations (2). * This study was supported by a grant No. 81-07-01 from the Kuwait Foundation for the Advancement of Sciences (KFAS). ' Associate Professor, Department of Microbiology, Faculty of Medicine, Kuwait University, Post Box No. 24923 Safat, Kuwait. All correspondence should be addressed to this author. 2 Associate Professor, Department of Paediatrics, Faculty of Medicine, Kuwait University. 3Chief Technical Officer, Faculty of Medicine, Kuwait University. 4Associate Professor, Department of Obstetrics and Gynae- cology, Faculty of Medicine, Kuwait University. s Associate Professor, Department of Community Medicine and Behavioural Sciences, Faculty of Medicine, Kuwait University. 6 Professor, Department of Obstetrics and Gynaecology, Faculty of Medicine, Kuwait University. 7 Professor, Department of Medicine, Faculty of Medicine, Kuwait University. 8 Project Co-ordinator, Department of Microbiology, Faculty of Medicine, Kuwait University. The present study attempts to define the role of several viruses and toxoplasma in causing congenital infection and their possible association with mal- formations among the offspring of Arab women in Kuwait. MATERIALS AND METHODS Design of the study Infants born with an elevated IgM in cord blood (> 16.4 mg/100 ml) were selected at birth and fol- lowed up serologically and clinically for the first year of life. Evidence of specific intrauterine infection was obtained by investigating persistent immune responses to RV, CMV, HSV, varicella-zoster virus, measles and mumps viruses, polyoma-BK virus, enteroviruses (polio-coxsackie-echo), and toxo- plasma for a period longer than the duration of maternal antibodies in the infants' blood (normally 6 months). Infants demonstrating an elevated cord blood 1gM and a persistent immune response were then investigated for the presence of specific IgM in the cord blood and/or in day- I neonatal serum to pro- vide further evidence of congenital infection and to exclude an early postnatal infection, especially in cases of CMV infection. Data were analysed using standard statistical methods, the association between two variables being assessed by the x2 test and the test of significance between two proportions. 4579 745- W. AL-NAKIB ET AL. A total of 795 women delivering at the Kuwait Maternity Hospital (which is the main maternity hospital in the country) were investigated over a period of 6 months during 1979. Cord blood and mothers' blood were collected and sent to the virus laboratory at the Medical School, University of Kuwait, for estimation of the total IgM. This was qualitatively estimated using a latex agglutination procedure (Rapitex IgM, Behringwerke, AG) and calibrated to give a positive test at levels of > 16.4 mg/100 ml. In cases that were positive, neonatal blood was collected and later investigated for an elevated IgM (> 16.4 mg/100 ml) in order to exclude maternal contamination of fetal blood which could give rise to false positive tests in the cord blood. Infants showing a positive IgM () 16.4 mg/100 ml) were then followed longitudinally for the first year of life. All the babies in the study underwent detailed physical and neurological examination between day 2 and day 7 after birth. A developmental assessment, based on the method described by Sheridan et al. (3), was carried out on all of them at 4-5 weeks, 5-6 months, 8-9 months, and 1 year of age. Blood was collected during each of these visits and serial speci- mens from each infant were investigated in parallel for antibody to all the infective agents mentioned above. Serology Sera were examined for rubella antibodies by a modified haemagglutination-inhibition (HAI) pro- cedure using trypsinized human 0 erythrocytes instead of day-old chick erythrocytes (4), while toxoplasma antibodies were determined using a passive-haemagglutination (PHA) procedure (5). The immune response to other viral agents was investigated using a modified, sensitive complement- fixation (CF) procedure in which serum-antigen mixtures were incubated overnight at 4 OC (6). The sensitivity of this CF test for CMV and HSV had earlier been evaluated against a sensitive PHA procedure modified in our laboratory (7). Antibody titres of > 1:8 in the HAI and CF tests were con- sidered positive, whereas titres of > 1:16 in the PHA test were considered positive for toxoplasma. Standard control sera with high positive titres ( > 1:128 for HAI and PHA, and 1:64 for CF), as well as sera with low (1:16) and negative (< 1: 8) titres were included in each test. Specific IgM in the cord blood in day- I neonatal sera was investigated by serum fractionation on sucrose density gradients (SDG) fol- lowed by a long-incubation HAI test for rubella (8) or by immunofluorescence for cytomegalovirus (9) and toxoplasma (10). SDG fractionation prior to specific IgM detection was considered to be imperative in order to reduce the likelihood of rheumatoid factor (RF) interfering with the test (11); RF in the IgM fractions was checked by employing a commercially available latex agglutination procedure. Quantitative estimation of the total IgM was conducted using sensitive S-partigen plates (Behringwerke, AG). RESULTS Of the 795 infants investigated for an elevated IgM () 16.4 mg/100 ml) in the cord blood, 72 (9.1%) were found to be positive, 65 (90.3%) of whom were later confirmed to have total IgM levels of over 16.4 mg/100 ml by radial immunodiffusion (RID). Of these 65 cases, 42 completed a clinical and sero- logical follow-up for a period of one year after birth, 23 (54.8%) of whom did not show any congenital defect or neonatal morbidity, or an immune response persisting beyond the normal duration of maternal antibodies (i.e., 6 months after birth) to any of the Table 1. Number of cases with a congenital defect and/or infection among babies born with an elevated IgM in the cord blood No. of cases Level of IgM in cord Total blood (mg/100 ml) With no congenital defect With congenital With serological With both congenital defect or serological defect only evidence of and serological evidence evidence of infection infection' of infection 16.4-30 16 (38)" 2 (4.8) 4 (9.5) 0 (0) 22 > 30 7 (16.6) 1 (2.4) 8 (19) 4 (9.5) 20 Total 23 (54.8) 3'(7.1) 12 (28.6) 4 (9.5) 42 Indicated by persistent immune responses. " Figures in parentheses are percentages. Three more cases had congenital defects and 2 further cases were underweight for the age, but we could not complete the sero- logical follow-up studies. In addition, we had one case of presumed cot death; no autopsy was available. 746 ELEVATED IGM IN CORD BLOOD IN KUWAIT Table 2. Data on 16 cases with serological evidence of intra-uterine viral infections Total IgM in Type of persistent Virus-specific IgM in cord blood immune responsea cord blood and/or day-1 (mg/100 ml) neonatal blood 50 50 36 50 39 50 34 27 40.25 50 29.5 25.5 21.5 CMV CMV CMV CMV CMV RV CMV CMV CMV RV CMV CMV CMV CMV CMV CMV 25.5 33.5 NDb ND ND ND ND ND ND Total number of cases with persistent immune responses to RV: 2/42 (4.8%) Total number of cases with persistent immune responses to CMV: 14/42 (33.3%) Total number of CMV cases tested and confirmed by virus- specific IgM: 8/9 (88.9%) a CMV = cytomegalovirus; RV = rubella virus. b ND = not done owing to lack of sera. infective agents studied. Three (7.1 7o) infants, how- ever, had congenital defects without any evidence of a persisting serological response. Twelve (28.6%) infants showed persistent responses during the whole period of the study and a further 4 infants (9.5%) had both serological evidence of infection and a congeni- tal abnormality (Table 1). As shown in Table 2, two (4.8%o) of these responses were due to rubella and 14 (33.3%o) were due to CMV. Eight of 9 cases (88.9%7o) were confirmed to have had a CMV infection by the demonstration of CMV-specific IgM in the cord blood and/or day-i neonatal serum. However, RV- specific IgM could not be investigated in either the cord blood or in day-I neonatal serum because of lack of sera. There was a significant association (P < 0.01) between the level of total IgM in the cord blood and the selection of cases with a persistent immune response. Fig. 1 shows the persistent immune responses (expressed as geometric mean titres) in the 2 cases of rubella and 14 cases of CMV, and the absence of such a response in the remaining 26 cases studied. The figure shows that for infants not showing a persistent response, most had lost their maternal antibodies by 5-6 months for RV and CMV, and by 4-5 weeks for HSV, measles, mumps, picornaviruses (polio- coxsackie-echo), varicella-zoster virus, polyoma-BK virus, and toxoplasma. In contrast, in the infected cases, whether due to rubella virus or CMV, the im- mune reponses not only persisted but the titres con- tinued to rise at 5-6 months and were still rising at 8-9 months of life. Table 3 shows some clinical abnormalities detected in 12 cases (18.5%) among the 65 infants with an elevated cord IgM during the period of the study. Of particular significance, perhaps, are the 5 infants who had a systolic murmur, which was thought to be due to a ventricular septal defect. Only 4 (33.3%o) of these cases had a persistent immune response; 3 (2 with a systolic murmur and 1 with harelip) were due to CMV infection and 1 (with a systolic murmur) to rubella. Thus, of the 795 cases initially screened, 12 (1.5%) Table 3. Types of clinical abnormality detected within 9-12 months of birth in 12 infantsa who were born with an elevated IgM (> 16.4 mg/100 ml) in the cord blood (confirmed by radial immunodiffusion) Clinical abnormality Total IgM Persistence in each patient concentration of antibodies (mg/100 ml) Systolic murmur" Para-umbilical hernia Systolic murmur" Umbilical hernia Harelip Systolic murmur" Coarctation of aorta and cleft palate Underweight for age Underweight for age Murmur" Underweight for age (< 2500 g) and inguinal hernia Murmur" 36 + (CMV)' 25.5 - 39 + (CMV) 35 - 40.25 + (CMV) > 50 + (RV)' 32.5 26.25 21.5 21 ND' ND ND ND 25.5 33.5 Proportion of cases with abnormality out of the total with an elevated IgM in cord blood: 12/65 (18.5%). Proportion of cases with abnormality who also showed serological evidence of intrauterine viral infection: 4/12 (33.3%) A further baby had a cot death at the age of 3 months; rmaterial for autopsy was not available. b Presumably due to a ventricular septal defect. CMV = cytomegalovirus; RV = rubella virus; ND = test not done. 747 W. AL-NAKIB ET AL. GMT OF CASES SHOWING PERSISTING ANTIBODIES TO RV 2/42 (4.8%) RV CMV HSV MEASLES MUMPS PICORNA V-Z POLYOMA TOXOPLASMA GMT OF CASES SHOWING PERSISTING ANTIBODIES TO CMV 14/42 (33.3%) RV CMV HSV MEASLES MUMPS PICORNA V-Z POLYOMA TOXOPLASMA GMT OF CASES NOT SHOWING PERSISTING ANTIBODIES 26/42 (61.9%) MCDWABMC DWAB MCDWABMCDWABM CDWABMC DWABMC DWAB MC DWABMCDWAB Fig. 1. Immune responses (indicated by geometric mean titres (GMT)) to viral and toxoplasmal infections (see text) among infants born with an elevated IgM in cord blood during the first 9 months of life. (M = mother's blood; C = cord blood; D = on day 1; W = at 4-5 weeks; A = at 5-6 months; B = at 8-9 months). were found to have a clinical abnormality, while 14 (1.80/o) had serological evidence of a CMV infection and 2 (0.25%) of rubella virus infection. DISCUSSION Significantly higher levels of total IgM in the cord blood were detected among the 14 cases with sero- logical evidence of intrauterine CMV infection (average, 39.5 mg/100 ml) when compared with those who showed no evidence of such infection (P< 0.01); levels of over 20 mg/100 ml have generally been accepted as indicating intrauterine infection (12). Furthermore, in contrast to the 26 cases who lost their maternal CMV antibodies at 5-6 months after birth, these 14 cases showed a persistent and rising CMV antibody level even at 8-9 months of life. This type of serological picture has been accepted as indicative of intrauterine infection provided it was not complicated by an early post-natal infection which is especially common in CMV infections (12). Using the indirect immunofluorescence procedure, CMV-specific IgM was demonstrated in the cord blood and/or in day-i neonatal serum in 88.9% of cases who had a persistent CMV immune response indicating that infection was not due to early post- natal infection. The specificity of our indirect immunofluorescence procedure for CMV-specific IgM was intensified by our routine sera fractionation on sucrose density gradients prior to immunofluorescence testing. Such a procedure has been shown to reduce the possibility of a false negative result due to competition of specific IgG with specific IgM for the same antigen- binding sites or false positives due to the presence of rheumatoid factor (11). Furthermore, our CMV- specific IgM fractions were shown to be free of 200 175 150 125 100 75 50 25 (U4 - co g 100 .0 75 E 50 0 '0 250 100 75 50 25 748 ELEVATED IGM IN CORD BLOOD IN KUWAIT rheumatoid factor when tested by latex agglutination. It is noteworthy that these 14 cases account for 1.8% of infections among our randomly selected 795 mothers. This figure is in line with an estimated range of 0.5-2.5% infection rate for CMV in a given popu- lation (12, 13). It must be emphasized that our data regarding CMV infection would have undoubtedly been strengthened had we provided further evidence of such infections by virus isolation from the urine of these infants during the first week of life. This, how- ever, was not in the original plan but is being investi- gated in the second phase of the study. Evidence implicating the 2 rubella cases was similar to that obtained for CMV cases except that early post- natal infection by RV is uncommon (12). In a single study in 1967, the rate of 'major' con- genital abnormalities detected in Kuwait during the first week of life was estimated to be 2.2%o of all live births (1). In recent unpublished studies in two of the major neonatology units in the country, the rate of congenital anomalies detected within the first few days of life was found to be 1.3%o (39 of 3000 new- borns studied) and 1.49% (122 of 8190 newborns studied) (L. V. Deverajan, personal communication, 1983). The latex agglutination procedure employed in our investigation was not only useful for a rapid (5 minutes) screening of infants with elevated cord blood IgM, and for selection of a relatively high pro- portion of cases who were later shown to have sero- logical evidence of intrauterine infection, but it also selected 12 (18.5%7o) out of 65 cases with clinical abnormalities, a proportion not normally seen in our daily practice for a given group. In addition, some 'minor' or 'silent' anomalies might not have been detected and these may show up later in life. It may be of interest to note that among our infants with elevated cord blood IgM ( > 16.4 mg/100 ml), the majority of cases presenting with clinical abnormalities generally had high levels of cord blood IgM (range 21 to >,50 mg/100 ml; mean 32 mg/100 ml). Similarly, those with serological evidence of congenital infection also generally had a high level of cord blood IgM (range, 21.5-71 mg/ 100 ml; mean, 39.5 mg/100 ml). This is in contrast to a mean IgM level of approxiruately 21 mg/100 ml, as detected by radial immunodiffusion in the cord blood of unselected normal infants in Kuwait. These obser- vations are further strengthened by our recent find- ings, employing the more sensitive laser nephelo- metry, that among a group of 241 infants screened for elevated IgM, only 1 (0.8807o) out of 114 infants with cord blood IgM levels of < 25 mg/100 ml had an abnormality (low birth weight). In contrast, 10 (7.88Gb) out of 127 infants with abnormalities (8 were underweight for age, 1 had a systolic murmur, and 1 had a cavernous haemangioma) had cord blood IgM levels of > 25 mg/100 ml (majority of these cases had levels above 35 mg/100 ml). These observations, therefore, suggest that an elevated IgM in the cord blood is a suitable criterion for selecting a high pro- portion of infants with various abnormalities includ- ing low birth weight. Nevertheless, a more accurate evaluation of this procedure for selecting such affected infants would have necessitated the follow- up of an equal number of a matched control group without an elevation of total IgM in the cord blood. This, however, we did not think was ethical and was therefore not done. Recently, evidence has been presented indicating that a combined testing of elevated total IgM, rheumatoid factor, and specific CMV-specific IgM in the cord blood appears to have significant prognostic potential for those babies who are initially asymptomatic (14). It is of interest that 5 (41.7%) out of our 12 cases with clinical abnormalities had a heart murmur; the significance of this is difficult to establish at present except that two of them were associated with CMV and one with RV. A further case with harelip was serologically associated with CMV. Four cases were underweight for age; two had other complications accompanying the low birth weight. Only harelip has previously been reported to be associated with a congenital CMV infection (13), since cardiovascular abnormalities have rarely been documented in such a situation. In our study we have presented serological evidence of congenital CMV infection in at least 14 (1.8%) of 795 cases tested, some of whom also had clinical abnormalities. This finding is noteworthy since about 95% of adult women in Kuwait possess antibody to CMV (1). It therefore remains to be seen whether reinfection, or more likely a reactivation of latent CMV infection, in our population is responsible for the development of congenital CMV infection and possibly fetal malformations. These results emphasize the need for further studies to define the role ofCMV infection during pregnancy among women in Kuwait. This infection together with rubella needs to be carefully monitored among this population in order that infants at risk can be identified early in life and therapeutic help may be given at the appropriate time. 749 750 W. AL-NAKIB ET AL. RESUME ETUDES SEROLOGIQUES ET CLINIQUES PROSPECTIVES CHEZ DES NOURRISSONS DU KOWEIT PRESENTANT UNE ELEVATION DES IgM DANS LE SANG DU CORDON L'6l6vation des IgM (> 16,4 mg/100 ml) a ete recherch6e dans le sang du cordon ombilical chez 795 nourrissons. Soixante-douze d'entre eux (soit 9,1%) presentaient une telle elevation, confirm6e par la suite en immunodiffusion radiale dans 65 cas (90,3%). Parmi ces 65 cas, 42 ont acheve le suivi clinique et serologique d'un an apres la naissance; dans ce groupe, 23 (54,8%) n'ont manifeste ni anomalie congenitale, ni morbidite neonatale, ni reponse immunitaire persistant au-dela de la dur6e de vie normale des anticorps maternels (soit environ 6 mois apres la naissance) diriges contre le toxoplasme ou l'un quelconque des virus du groupe etudie (virus rubeoleux, cytomegalovirus, virus herpetique alpha types 1 et 2, virus de la varicelle et du zona, virus rougeoleux et virus ourlien, virus BK et virus du polyome, enterovirus (poliovirus, virus Coxsackie, virus ECHO). Sur ces 42 nourrissons, 3 (7,1 %) presentaient toutefois des anomalies congenitales sans signe serologique d'une reponse immunitaire persistante, 12 (28,6%) ont manifeste pendant toute la duree de l'etude une reponse immunitaire persis- tante et 4 autres (9,5%) etaient porteurs A la fois de signes serologiques d'infection et d'une anomalie congenitale. Deux de ces reponses en anticorps (4,8%) etaient dirigees contre le virus rubeoleux et 14 (33,3%) contre le cyto- megalovirus (CMV). L'infection A cytomegalovirus a e confirmee chez 8 des 9 cas etudies par la mise en evidence d'IgM specifiques anti-CMV dans le sang du cordon et/ou dans le serum du nouveau-ne preleve au premier jour. On a observe une association significative (P <0,01) entre le taux d'IgM totales dans le sang du cordon et les cas qui se sont reveles porteurs d'une reponse immunitaire durable. La serologie a mis en evidence une infection virale chez 4 des 12 cas (33,3%) porteurs d'une anomalie clinique, 3 corres- pondant au cytomegalovirus (2 souffles systoliques et 1 bec- de-lievre) et 1 au virus rubeoleux (souffle systolique). REFERENCES 1. AL-NAKIB, W. ET AL. Seroepidemiology of viral and toxoplasmal infection during pregnancy among Arab women of child-bearing age in Kuwait. International journal of epidemiology, 12: 220-223 (1983). 2. EL-ALFI, 0. S. ET AL. Congenital malformation in Kuwait. Journal of the Kuwait Medical Association, 2: 99-108 (1968). 3. SHERIDAN, M. D. Children's developmental progress from birth to five years. London, NFER Publishing Company, 1977. 4. AL-NAKIB, W. & LILLEY, H. Detection of rubella haemagglutination inhibition (HAI) and virus-specific IgM antibody using trypsin-treated human group 0 erythrocytes in the HAI test. Journal of clinical pathology, 31: 730-734 (1978). 5. THORBURN, H. & WILLIAMS, H. A stable haemag- glutinating antigen for detecting toxoplasma anti- bodies. Journal of clinical pathology, 25: 762-767 (1972). 6. GRIST, N. R. ET AL. Complement fixation test. In: Diagnostic methods in clinical virology, 3rd ed., Oxford, Blackwells, 1979, pp. 95-115. 7. AL-NAKIB, W. A modified passive haemagglutination technique for the detection of cytomegalovirus and herpes simplex virus antibodies: application in virus- specific IgM diagnosis. Journal of medical virology, 5: 287-293 (1980). 8. AL-NAKIB, W. ET AL. Rubella-specific IgM and a new inhibitor. British medicaljournal, 3: 579 (1974). 9. HANSHAW, J. B. ET AL. Fluorescent antibody test for cytomegalovirus macroglobulin. New England journal of medicine, 279: 566-570 (1968). 10. CHESSUM, B. S. Examination of sera for toxoplasmosis antibody using immunofluorescence. Journal of medical laboratory technology, 27: 49-54 (1970). 11. SCHMITZ, H. ET AL. Detection of IgM antibodies to cytomegalovirus (CMV) using an enzyme-labelled anti- gen (ELA). Journal of general virology, 50: 59-68 (1980). 12. HANSHAW, J. B. & DUDGEON, J. A. Laboratory diagnosis. In: Viral diseases of thefetus and newborn. Philadelphia, Saunders Company, 1978, pp. 301-376. 13. STAGNO, S. ET AL. Perinatal infections and maldevelop- ment. Birth defects, 17: 31-50 (1981). 14. GRIFFITHS, P. D. ET AL. Congenital cytomegalovirus infection. Diagnostic and prognostic significance of the detection of specific immunoglobulin M antibodies in cord serum. Journal ofpediatrics, 69: 544-549 (1982).