Association of clinical presentation, laboratory findings, and virus serotypes with the presence of meningitis in hospitalized infants with enterovirus infection

ORIGINAL ARTICLES

Association of clinical presentation, laboratory findings, and virus serotypes with the presence of meningitis in hospitalized infants with enterovirus infection Ron D a g a n , MD, Jerri A. Jenista, MD,* a n d Marilyn A. M e n e g u s , PhD From the Departments of Pediatrics, Microbiology, and Pathology, The Universityof Rochester Medical Center, Rochester, New York

One hundred eight hospitalized infants with enteroviral infections were studied to determine the association of clinical presentation, laboratory findings, and virus serotypes with the presence of meningitis. Of 108 infants, 55 (51%) had meningitis. Clinical manifestations on admission did not distinguish between infants with and those without meningitis. Echoviruses 30 and 11 and coxsackie virus B were frequently associated with meningitis (34/38; 90%) whereas echoviruses 18, 24, and 25 were not (5/35; 4%). The virus isolation rate was directly proportional to the number of leukocytes in cerebrospinal fluid: 5 of 58 (9%) when up to 9 cells/mm 3 were found, 10 of 21 (48%) when 10 to 99/mm 3 cells were found, and 25 of 29 (86%) when ~100 cells/mm 3 were found. Meningitis is often unsuspected in children hospitalized with enterovirus infection. The frequency of meningitis among hospitalized infants is serotype d e p e n d e n t and is most frequently, but not exclusively, found with pleocytosis of the cerebrospinal fluid. (J PEDIATR1988;113:975-8)

Infections caused by nonpolio enteroviruses are common among infants~6; those hospitalized with such infections usually have meningitis or an acute febrile illness without meningitis. The morbidity associated with meningitis caused by enteroviruses has been the subject of many reports, but few have included infants with nonmeningitis febrile illness and none has reported the relative frequency of each presentation among hospitalized infants or contrasted the two in terms of clinical and laboratory variables or etiologic factors. Distinguishing between meningitis and Dr. Dagan is the recipient of the University of the RochesterIsraeli Pediatric Scholar Exchange Award. Submitted for publication Feb. 29, 1988; accepted June 16, 1988. Reprint requests: Ron Dagan, MD, Pediatric Infectious Disease Unit, Soroka Medical Center, P.O. Box 151, Beer-Sheva 8410, Israel. *Now at the Department of Pediatrics and Communicable Diseases, C. S. Mott Children's Hospital, Ann Arbor, Michigan.

a sepsislike picture without meningitis may have an impact on both hospital management4,7 and long term follow-up, because meningitis in very young infants is rarely associated with neurologic impairment.8.H CSF

Cerebrospinalfluid

This study was designed to determine (1) whether clinical findings differ in enterovirus-infected infants with and without meningitis,(2) whether the frequency of meningitis varies relative to the infecting serotype, and (3) whether there is a relationship between the cerebrospinal fluid leukocyte count and the recovery of virus from the CSF. METHODS Virus cultures were obtained from infants <1 year of age hospitalized between July 1, 1982, and October 31,

975

976

Dagan, Jenista, and M e n e g u s

The Journal o f Pediatrics December 1988

T a b l e I. Frequency of CSF pleocytosis and positive CSF

virus cultures in 108 infants <1 year of age with enterovirus infection

T a b l e II. Clinical presentation of 108 hospitalized infants with enterovirus disease Infants with meningitis (n = 55)

No. ( % ) of infants with CSF

No. of infected infants

WBC + Virus +

WBC + Virus -

108

35 (32)

15 (14)

WBC Virus +

5 (5)

WBC Virus -

53 (49)

WBC +, ~ 10 white bloodcells/ram3of CSF; WBC-, <10 whiteblood

celts/mm3of CSF.

1983, at the Strong Memorial Hospital with at least one of the following clinical presentations: acute fever of unknown origin, suspected sepsis, meningitis, and meningoencephalitis. This population represents all infants <3 months of age who were hospitalized during the abovementioned period for suspected sepsis and the infants < 1 year of age who were hospitalized for suspected meningitis. Fecal, throat, blood, and CSF specimens were obtained from each infant and processed as described previously,n Enterovirus disease was defined if enterovirus was isolated from at least one site, no other pathogens were isolated, and no other diagnosis was established. Details of the history and physical examination were collected from the parents and the hospital chart at the time of admission. Laboratory evaluations were recorded from the computerized laboratory reports. Pleocytosis was defined as the presence of -> 10 leukocytes/mm 3 in the CSF. Infants whose CSF contained >I000 erythrocytes/mm3 were excluded. In addition, infants with positive blood culture for an enterovirus were excluded if the CSF contained > 100 erythrocytes/mm3, to exclude the possibility of CSF contamination with viruspositive blood. Meningitis was defined as the presence of pleoeytosis or enterovirus in the CSF. Chi-square tests with Yates correction and Student t tests were used for the data analysis. A p value of -->0.05 was considered significant except for multiple comparisons, for which a p value of <0.01 was considered significant. RESULTS Enterovirus was isolated from 153 patients. Eight patients were excluded because CSF was not obtained, and 37 were excluded because CSF could not be evaluated because of the presence of erythrocytes. Thus the final study group consisted of 108 infants. The median age was 42 days (range 4 days to 1 year). All infants were admitted for suspected sepsis or suspected meningitis. Enterovirus was isolated most frequently from stool (93/108; 86%),

Male Age <3 mo Fever >~38.3~ C Mean temperature Irritability Lethargy Poor feeding Loose stools Vomiting Rash URI

Infants without meningitis (n = 53)

No.

%

No.

%

29 44 53 39.0 40 30 25 23 13 13 9

53 80 96 38.1-40.2 72 55 45 42 24 24 16

25 46 49 39.0 48 26 24 20 16 10 8

47 87 92 37.7-40.5 91 49 45 38 30 19 15

URL Upper respiratorytract infection.

followed by throat (62/108; 57%), CSF (40/108; 39%) and blood (36/108; 33%). Of the 108 study infants, 55 (51%) had meningitis (Table I). The clinical presentation of the infants with and without meningitis is shown in Table II. None of the variables in the table differed significantly between the groups with and without meningitis. A bulging fontanelle or meningeal signs were not present in any of the 55 infants with meningitis. The frequency of meningitis in infants infected with the various enterovirus serotypes and the range of CSF leukocyte counts are presented in Table III. Infections with echoviruses 30 and 11 and the type B coxsackie viruses were associated with a high frequency of meningitis (34/38; 90%), whereas echoviruses 18, 24, and 25 were significantly less frequently associated with meningitis (5/35; 14%) (p <0.001). The CSF cell counts among infected infants ranged up to 2262/#1. Of the 55 infants with meningitis, 29 (53%) had ->100 leukocytes3 and 9 (16%) had >500 leukocytes/mm3. A pleocytosis of >100/ mm 3 was noted in 23 of 34 cases (69%) of meningitis caused by echoviruses 30 and 11 or type B coxsackie viruses, in comparison with l to 7 cases (14%) of meningitis involving the low-frequency meningitis enterovirus serotypes (p <0.01). Tables I and IV present the relationship between CSF leukocyte counts and the frequency of virus recovery from CSF. Virus isolation was significantly more frequent when more ceils were present. However, in 9% of infected infants without pleocytosis, enterovirus was isolated from CSF. Of the 40 culture-positive CSF specimens, 5 (12%) did not show pleocytosis. The CSF glucose levels were lower with

Volume 113 Number 6

Enteroviral meningitis in infants

977

Table III. Frequency of meningitis in infants with various enterovirus serotypes and range of CSF leukocyte counts Cases of meningitis

No. of

Leukocytes/mm 3 in CSF

Enterovirus

infected infants

No.

%

Range

Median

Echovirus 30 Coxsackievirus type B* Echovirus 11 Echovirus 25 Echovirus 24 Echovirus 18 Othersi" Total

10 20 8 8 17 10 _35 108

10 18 6 2 4 1 14 55

100 90 75 25 24 10 34 51

13-858 3-2262 1-397 0-415 1-50 0-8 0-620 0-2262

355 288 10 3 3 2 7

*CoxsackievirusesB2(2), B3(2), B4(3), B5(13). tEchoviruses 5(2), 6(1), 7(1), 9(4), 12(1), 14(2), 16(1), 17(2), 20(2), 21(1), 22(2), 23(3), coxsackievirusA9(3), untypedA(3), enterovirus71(1), untyped EV(6). meningitis than without meningitis (meningitis present: 49.7 + 12.2 mg/dl [2.76 + 0.68 mmol/L], range 25 to 111 mg/dl [1.39 to 6.15 mmol/L]; meningitis absent: 54.6 _+ 10.6 mg/dl (3.03 + 0.59 mmol/L], range 39 to 94 mg/dl [2.16 to 5.21 m m o l / L ] ; p <0.025). The CSF protein levels were higher with meningitis than without meningitis (with meningitis: 54.9 + 22.7 mg/dl [range 21 to 117]; withoug meningitis: 41.1 _+ 14.4 mg/dl [range 12 to 86];p <0.001). DISCUSSION Our study demonstrates that in most instances, enterovirus-infected infants with and without meningitis are clinically indistinguishable. The frequency of meningitis varied considerably with the infecting serotype, and the rate varied directly with the CSF leukocyte count. However, even in the absence of pleocytosis, 9% of CSF specimens yielded virus. Meningitis was not suspected clinically in most infants in whom it was found. In our study the diagnosis of meningitis was based on a single lumbar puncture performed on the day of admission. The diagnosis was established by the presence of pleocytosis, positive CSF culture, or both. The CSF cellular reaction to infection is a dynamic process, and it is possible that in some Of the infants who were defined as having meningitis, on the basis only of positive virus culture, pleocytosis developed later. Moreover, it is also possible that with single lumbar puncture, many cases of meningitis in enterovirus-infected infants are missed. Therefore it is possible that the true frequency of meningitis among symptomatic enterovirus-infected infants is higher than the figure of 51% found in our study. Meningitis was more frequent in infants infected with certain enterovirus serotypes. It is appealing to speculate

T a b l e IV. CSF leukocyte counts and frequency of virus

isolation Infants with virus isolated from CSF

No. of leukocytes/mm ~ in C$F

No. of infants

No.

%

0-9 10-99 >100

58 21 29

5 10 25

9 48 86

that these serotypes are more neurotropic than others. Coxsackieviruses type B were the most common identified cause of aseptic meningitis at Waiter Reed hospital from 1953 to 1963,"' ~4whereas the Centers for Disease Control reported echoviruses as the most frequent isolate from meningitis cases in the United States from 1970 to 1979.15 However, the frequency of meningitis reported in such surveys tends to reflect the prevalence of serotypes in the community. Even if an enterovirus does not cause meningitis at a high frequency, it can be the most common cause of meningitis if its prevalence is very high. In our study, although the numbers were small, denominator data for infants ill enough to be hospitalized are provided, and the relative frequency of meningitis with different enterovirus serotype infections could be calculated. Some enterovirus serotype (coxsackie virus B, echoviruses 11 and 30) were associated with a greater frequency of meningitis than others (echoviruses 18, 24, and 25). However, frequent antigenic changes found among naturally occurring enterovirus16may also result in changes in virulence and in tissue tropism.

97 8

Dagan,Jenista, and Menegus

In our series, the virus isolation rate increased with increasing C S F leukocyte counts. However, it seems advisable to perform virus cultures regardless of the presence or absence of pleocytosis, particularly during periods of high enterovirus prevalence. Meningitis is often unsuspected in very young infants hospitalized with enterovirus infection. We have found that the frequency of enterovirus meningitis among hospitalized infants is serotype dependent and is most frequently, but not exclusively, found with C S F pleocytosis. REFERENCES

1. Melnick JL. Viral infections of humans: epidemiology and control. In: Evans AS, eds. Enteroviruses. New York: Plenum Medical, 1982:187-249. 2. Dagan R, Jenista JA, Menegus MA. Clinical, epidemiological and laboratory aspects of enterovirus infection in young infants. In: De La Maza LM, Peterson EM, eds. Medical virology; vol IV. Hillsdale, N.J.: Lawrence Erlbaum Associates, New Jersey, 1985. 3. Centers for Disease Control. Aseptic meningitis Surveillance Summary 1970-1979. Washington, D.C.: U.S. Department of Health and Human Services, Nov 1981. 4. Chonmaitree T, Menegus MA, Powell KR. The critical relevance of "CSF viral culture": a two-year experience with aseptic meningitis in Rochester, N.Y. JAMA 1982; 247:18437. 5. Barrett M J, Strikas RA, Rogers MF, Rabkin C, Alexander WJ. Aseptic meningitis in Jefferson County, Alabama [Abstract]. Pediatric Res 1984;18:180A. 6. Cherry JD. Non-polio enteroviruses. In: Feigin RD, Cherry

The Journal of Pediatrics December 1988

7.

8.

9.

10,

11.

12.

13.

14.

15.

16.

JD, eds. Textbook of pediatric infectious diseases. Philadelphia: WB Saunders, 1981:135-6. Jenista JA, Prather SA, Powell KR, Menegus MA. Enterovirus study group: virus cultures in the reduction of neonatal morbidity [Abstract]. Pediatric Res 1983;17:224A. Willfert CM, Thompson R J, Sunderr TR, O'Quinn A, Zeller J, Blacharsh J. Longitudinal assessment of children with enteroviral meningitis during the first three months of life. Pediatrics 1981 ;67:811-5. Sells CH, Carpenter RL, Ray GC. Sequelae of centralnervous system enterovirus infections. N Engl J Med 1975;293:1-4. Jenista JA, Dalzell LE, Davidson PW, Menegus MA. Outcome studies of neonatal enterovirus infection [Abstract]. Pediatric Res 1984;18:230A. Bergman I, Painter M J, Wald ER, Chiponis D, Holland AL, Taylor HG. Outcome of children with enterovirus meningitis during the first year of life. J PEDIATR 1987;110:705-9. Dagan R, Jenista JA, Prather SL, Powell KR, Menegus MA. Viremia in hospitalized children with enterovirus infections. J PEDIATR 1985;106:397-401. Meyer HM Jr, Johnson RT, Crawford IP, Dascomb HE, Rogers NG. Central nervous system syndromes of "viral" etiology: a study of 713 cases. Am J Med 1960;29:334-47. Buescher EL, Artenstein M J, Olson LC. Central nervous system infection of viral etiology: the changing patterns. Res Publ Assoc Res Nerv Merit Dis 1968;44:147-63. Moore M, (Centers for Disease Control). Enteroviral disease in the United States, 1970-1979. J Infect Dis 1982;146:1038. Prabhaker BS, Menegus MA, Notkins AL. Detection of conserved and non-conserved epitopes on coxsackievirus B4: frequency of antigenic changes. Virology 1984;143:302-6.