J Neurosurg: Pediatrics / Volume 13 / February 2014

J Neurosurg Pediatrics 13:189–191, 2014

189

©AANS, 2014

In developing countries, myelomeningoceles (MMCs) are rarely diagnosed prenatally. Most mothers receive no prenatal care, and their infants are often born at

home in unclean conditions. Infants born with MMCs may not be brought for medical care for several days or weeks because of local taboos against seeking medical care, inability to afford transportation to medical care, or unavailability of medical care.

By the time they present for care, infants are often febrile and their MMCs are often infected. Whether the infection evident on the neural placode extends to CSF underneath or extends upward into ventricular CSF could alter their management in terms of administration of an-tibiotics and timing of MMC closure.

We performed this study to evaluate CSF from lat-eral ventricles and MMCs at the time of presentation for medical care.

MethodsRecords were reviewed for 100 consecutive children

with MMCs who presented for treatment at the Kijabe Hospital. White blood cell (WBC) counts were deter-mined in peripheral blood specimens. CSF was aspirated from the lateral ventricles with a 22-gauge spinal needle and from the MMC sac below the neural placode. CSF was sampled from all intact MMC sacs that contained CSF. CSF was removed with a 23-gauge needle inserted into the caudal, insensate, end of the sac via a trajectory parallel to the dorsal plane of the patient’s back. WBC counts in CSF from both sites were determined by the laboratory. CSF from either site that contained 5 or more WBCs was cultured; that number was arbitrarily chosen, although previous CSF cultures with WBC cell counts lower than 5 cells/mm3 were virtually never positive. The progressively longer age categories chosen (< 1 week, 1 week–1 month, 1–3 months, and > 3 months) reflect the progressively smaller number of patients seen at each time interval.

Cerebrospinal fluid white blood cell counts in infants with myelomeningoceles

Clinical article

A. LeLAnd ALbright, M.d., SuSAn S. FerSon, C.P.n.P., And huMPhrey okeChi, M.b.Ch.b.Division of Neurosurgery, Department of Surgery, Kijabe Hospital, Kijabe, Kenya

Object. The authors undertook this study to determine white blood cell (WBC) counts in CSF obtained from lateral ventricles and myelomeningoceles (MMCs) in infants in a developing country at the time of their initial pre-sentation for medical evaluation.

Methods. CSF was aspirated from the lateral ventricles and from MMC sacs of 100 consecutive infants at Kijabe Hospital, Kijabe, Kenya. Peripheral blood WBC counts and CSF WBC counts were determined in the laboratory. CSF with WBC counts of 5 cells/mm3 or greater was cultured.

Results. The mean WBC count in ventricular CSF was 16 cells/mm3, with a median and mode of 0 cells/mm3. The mean WBC count of CSF in MMC sacs was 141 cells/mm3 (median 15 cells/mm3). No child had both a positive culture from ventricular CSF and a negative culture from MMC CSF. There was no correlation between age at pre-sentation and WBC counts in the MMCs. Infants younger than 8 days old were as likely to have high WBC counts in CSF from their MMC sacs as were older children; 7 of 12 infants with 500 WBCs or more in CSF from their MMCs were younger than 8 days old. Only 5 of 58 CSF specimens from MMC sacs with 5 or more WBCs/mm3 had posi-tive bacterial cultures, which may be a reflection of CSF specimen processing rather than of true culture negativity.

Conclusions. CSF from ventricular fluid of infants presenting with MMCs infrequently has high WBC counts, so infrequently that it does not need to be evaluated routinely. CSF in MMC sacs often has high WBC counts that sug-gest the presence of bacterial infection. In developing countries where culture reliability is questionable, intravenous administration of antibiotics before MMC closure for infants with high MMC WBC counts may diminish postopera-tive meningitis/ventriculitis.(http://thejns.org/doi/abs/10.3171/2013.11.PEDS13196)

key WordS      •      CSF analysis      •      myelomeningocele      •      infant      •      congenital

Abbreviations used in this paper: MMC = myelomeningocele; WBC = white blood cell.

A. L. Albright, S. S. Ferson, and H. Okechi

190 J Neurosurg: Pediatrics / Volume 13 / February 2014

Data correlations were evaluated by Pearson rank tests. The study was approved by the Kijabe Hospital in-stitutional review board.

ResultsThe patients’ ages ranged from less than 1 day (pre-

sentation on the day of birth) to 6.5 months; 51% were admitted within the 1st week of life, 27% between 1 week and 1 month of age, 17% between 1 and 3 months of age, and 5% when older than 3 months.

The WBC counts of peripheral blood did not corre-late with age (p = 0.94). The mean WBC counts in chil-dren 1 week old or less were 10,630 cells/mm3 (range 4100–27,900 cells/mm3). The counts were over 15,500 cells/mm3 in only 7 of 49 cases. Mean WBC counts in pe-ripheral blood of children 1 week–1 month, 1–3 months, and > 3 months of age were 11,570, 10,220, and 9800 cells/mm3, respectively. The mean WBC count of the 5 children with positive CSF cultures was 12,680 /mm3, within the normal range.

WBC counts in ventricular CSF ranged from 0 to 550 cells/mm3, with a mean of 16 cells/mm3 but a median and a mode of 0 cells/mm3. Two ventricular CSF specimens had positive cultures, one with 110 WBCs/mm3 that dem-onstrated growth of Escherichia coli and one with 400 WBCs/mm3 that demonstrated growth of Pseudomonas. These two children had WBC counts of 400 cells/mm3 and more than 1000 cells/mm3 in CSF from their MMCs, and cultures of the CSF were positive for the same organisms. No child had a positive culture from ventricular CSF but a negative culture from CSF obtained from the MMC.

WBC counts in MMC CSF ranged from 0 cells/mm3 to more than 1000 cells/mm3, with a mean of 141 cells/mm3 and a median of 15 cells/mm3. Thirty specimens had 0 WBCs, 42 had less than 5 WBCs/mm3, and 58 had 5 or more WBCs/mm3. Of the specimens with more than 5 WBCs/mm3, 22 had less than 100 cells/mm3, 12 had 100–500 WBCs/mm3, 5 had 500–1000 WBCs/mm3, and 7 had more than 1000 WBCs/mm3. As expected, neutro-phils accounted for the greatest proportion of cells in all of the elevated WBC counts. Only 5 MMC specimens had positive cultures: 2 demonstrated coagulase-negative Staphylococcus; 1, E. coli; 1, Enterobacter; and 1, Pseu-domonas. There was no correlation between the patients’ temperature on admission and WBC count in the CSF from the MMC: 62% of infants with 15 or more WBCs/mm3 in the CSF from their MMCs were afebrile, with temperatures of less than 38°C.

There was no correlation between patient age at admis-sion and WBC counts from the MMC CSF (p = 0.73); CSF from the MMC of a 1-day-old infant had 1520 WBCs/mm3 and a specimen from a 4-month-old infant had 0 WBCs. Conversely, many infants less than 1 week old had 0 WBCs and a 6-week-old infant had more than 1000 WBCs/mm3. Interestingly, 7 of 12 children with WBC counts of 500 cells/mm3 or more in CSF from their MMCs were 1 week old or younger, and 5 of 12 were only 1 or 2 days old.

The correlation between peripheral blood WBC counts and ventricular WBC counts was marginally sig-nificant (p = 0.027). There was a statistically strong cor-

relation between WBC counts in CSF from MMCs and ventricular fluid (p = 0.0005), because they both tended to be 0 at the same time. Of 13 children with > 20 WBCs/mm3 in their ventricular CSF, 6 had WBC counts of 35 cells/mm3 or less in CSF from their MMCs and 7 had counts of 175 WBCs/mm3 or more.

DiscussionAnalysis of CSF from MMC sacs is more likely to re-

veal high WBC counts than analysis of ventricular fluid is, and it is therefore probably a better indicator of the pres-ence of infection. Infants less than a week old were as like-ly to have high WBC counts in CSF from their MMC sacs as were older children. The finding of high WBC counts in CSF from MMC sacs at the time of admission altered our management: those infants were treated with intravenous antibiotics before the MMC was repaired. Treatment usu-ally began with ceftazidime (50 mg/kg/dose, administered every 12 hours for children younger than 8 days and every 8 hours for those 8 days old or older) and gentamicin (5 mg/kg/day in 1 dose) until culture results were available, though most cultures were sterile. When cultures were sterile, antibiotics were continued until the infants were afebrile for more than 48 hours, breast-feeding well, and no longer requiring supplemental oxygen.

Cultures were negative for 5 of 7 CSF specimens from MMCs with more than 1000 WBCs/mm3. The paucity of positive cultures is not an indication that the vast propor-tion of CSF specimens with high WBC counts were sterile but rather a reflection of the methodological problems in culture techniques in our hospital. Specimens may not be taken promptly to the laboratory, and once there, they may not be applied to culture plates for another 1–2 hours. After MMC closure, ventricular CSF was not routinely cultured, but rather was cultured only in cases in which the infants’ clinical signs suggested infection.

High WBC counts in children with MMCs who were less than a week old probably reflect the contaminated conditions of their birth. At least half of the children in the study were born at home, in homes with dirt floors and walls and without running water or ready availabil-ity of clean water. In those circumstances, MMCs are covered with whatever cloth is available, and it becomes soiled with urine and feces within hours.

We are aware of only 1 article comparing CSF from ventricles and MMC sacs. Kaplan et al.1 sampled CSF from 24 MMC sacs in 2 groups of infants—7 infants who were undergoing MMC repair within the first 24 hours of life and 17 who were undergoing repair after 24 hours. The WBC counts were below 10 cells/mm3 in ventricular and MMC sac fluid from infants less than 24 hours old. Our data were similar, with ventricular cell counts below 10 cells/mm3 in all infants younger than 24 hours old. In infants older than 24 hours, Kaplan et al. found no signifi-cant difference between the number of cells in CSF from ventricular fluid and CSF from the MMC sac, although cell counts and patient ages were not presented. Our data above document the disparity between analysis of CSF from the ventricles and that from MMC sacs.

Kaplan et al.1 also reported that in 4 patients with E.

J Neurosurg: Pediatrics / Volume 13 / February 2014

CSF WBC counts in infants with myelomeningoceles

191

coli infections in both ventricular CSF and MMC sacs, the number of cells in both sites decreased similarly in response to treatment. We did not repeat peripheral blood WBC counts during antibiotic therapy because of their cost in our environment. Kaplan et al. concluded that CSF can be readily obtained for analysis from MMC sacs, an observation confirmed in our experience, and that if the CSF has no evidence of infection, there is less infection if synchronous CSF shunting is performed, an observation our data cannot address.

We conclude that analysis of CSF from MMC sacs should be done in developing countries when possible if the results will alter treatment. We conclude that at the time of presentation of infants with MMCs, CSF from MMC sacs frequently contains high WBC counts sugges-tive of bacterial infection.

Because ventricular CSF usually has a low WBC count, it is appropriate to do initial ventricular taps in addition to MMC taps only in infants who appear to be septic when admitted. We prefer administration of intra-venous antibiotics until infants are afebrile and breast-feeding well before MMC closure, although no data are available to compare that method with operating on the MMC initially and then treating with antibiotics.

Acknowledgment

We gratefully acknowledge the statistical analyses performed

by Thomas D. Cook, Senior Scientist, Department of Biostatistics, University of Wisconsin Health Center.

Disclosure

The authors report no conflict of interest concerning the mate-rials or methods used in this study or the findings specified in the paper.

Author contributions to the study and manuscript preparation include the following. Conception and design: all authors. Acquisi-tion of data: all authors. Analysis and interpretation of data: Albright, Okechi. Drafting the article: Albright. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Albright.

Reference

1. Kaplan M, Ucler N, Bayrakli F, Duz B, Erol FS: Diagnosis of central nervous system infection by CSF sampling of the myelomeningocele sac as an alternative to ventricular tap. Neurocirugia (Astur) 21:228–231, 2010

Manuscript submitted April 19, 2013.Accepted November 14, 2013.Please include this information when citing this paper: published

online December 13, 2013; DOI: 10.3171/2013.11.PEDS13196.Address correspondence to: A. Leland Albright, M.D., Kijabe

Hospital, PO Box 20, Kijabe 00220, Kenya. email: l.albright@ neurosurgery.wisc.edu.