| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Imaging |
1 From the Departments of Child Neurology (M.S.v.d.K., G.V.) and Radiology (F.B.), Vrije Universiteit Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, the Netherlands; Departments of Clinical Epidemiology and Biostatistics (A.A.M.H.) and Medical Microbiology (J.F.L.W.), Academic Medical Center, Amsterdam, the Netherlands; and Diagnostic Laboratory for Infectious Diseases and Perinatal Screening, National Institute for Public Health, Bilthoven, the Netherlands (J.G.L.). Received November 6, 2002; revision requested January 15, 2003; final revision received June 10; accepted August 6. Address correspondence to M.S.v.d.K. (e-mail: ms.vanderknaap@vumc.nl).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: On the basis of findings in eight patients with documented congenital CMV infection, the authors developed MR imaging inclusion criteria, including multifocal lesions predominantly located in the deep parietal white matter. If gyral abnormalities were present, white matter lesions were either multifocal or diffuse. The criteria were applied to 152 patients with static leukoencephalopathy of unknown etiology. Guthrie cards for 22 of the 43 patients fulfilling the MR imaging criteria, 20 patients not fulfilling them, and 300 control subjects were analyzed. Fisher exact testing was used to evaluate the association between MR imaging characteristics and CMV status, and backward elimination linear discriminant analysis was used to identify MR imaging characteristics predictive of CMV infection in addition to the initial criteria.
RESULTS: PCR test results were positive in 12 of 22 patients suspected of having congenital CMV infection, in no patient not suspected of having infection (P < .001), and in two of 300 control subjects (negative predictive value [NPV] of MR imaging criteria, 100% [95% CI: 83%, 100%]; positive predictive value [PPV], 55% [95% CI: 32%, 76%]). The most important additional MR imaging finding predicting a positive PCR result was abnormality of the anterior part of the temporal lobe, including abnormal white matter, cysts, and enlargement of inferior horns. Including this finding in the MR imaging criteria enhanced the PPV (89%; 95% CI: 52%, 99%) at the expense of the NPV (88%; 95% CI: 72%, 97%).
CONCLUSION: In patients with static encephalopathy, an MR imaging pattern of multifocal lesions predominantly involving deep parietal white matter, with or without gyral abnormalities, is predictive of congenital CMV infection. When gyral abnormalities are present, leukoencephalopathy may also be diffuse. The presence of abnormalities in the anterior part of the temporal lobe increases the likelihood that CMV infection is present.
© RSNA, 2004
Index terms: Brain, diseases, 10.8721 • Brain, MR, 10.12141 • Brain, white matter, 10.2066 • Infants, central nervous system, 10.8721 • Magnetic resonance (MR), in infants and children, 10.12141 • Viruses, 10.2066
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It has been estimated that, despite these efforts, at least 50% of cases remain without a specific diagnosis (1). The battery of diagnostic tests is usually applied regardless of the pattern of abnormalities shown at MR imaging. However, MR imaging pattern recognition has proved to be a powerful diagnostic tool for childhood white matter disorders (1,3,4); use of MR imaging pattern recognition can reduce the number of possible diagnoses and, consequently, limit the number of diagnostic tests.
Recognition of distinct MR imaging patterns in patients with leukoencephalopathy of unknown cause can also be used to help define previously unknown disease entities (5–7). The underlying gene defects for two such disorders (megalencephalic leukoencephalopathy with subcortical cysts and leukoencephalopathy with vanishing white matter) have recently been identified (8–11), confirming the value and validity of this MR imaging–oriented approach. For these disorders, the diagnosis was initially MR imaging based, whereas DNA-based diagnosis is now available.
From a review of the literature (12–17) and from personal experience, we noted that cerebral white matter abnormalities of variable extent, with or without cerebral gyral abnormalities, may result from congenital cytomegalovirus (CMV) infection. The related encephalopathy is clinically static, in contrast to the classic leukodystrophies, which have a progressive course. In our patients with documented congenital CMV infection, we observed a particular pattern of white matter abnormalities at MR imaging. We observed a similar pattern in some patients who were referred to us because of a finding at MR imaging of leukoencephalopathy that had no identifiable cause, even after extensive metabolic testing was performed.
The problem with rendering a diagnosis of congenital CMV infection has been that it could not be confirmed after the neonatal period. In the absence of overt neonatal signs, infants with congenital CMV infection are not tested for the presence of the virus or immunoglobulin M antibodies within the appropriate time frame of 3 weeks. Those who develop neurologic sequelae usually present after the age of 6–9 months, and at that time it is no longer possible to confirm the diagnosis. More recently, a polymerase chain reaction (PCR) technique has been developed that reveals CMV DNA in blood spots collected on filter paper (Guthrie cards) during the neonatal period (18–20). In many countries, Guthrie cards are used for neonatal screening tests and are stored for a variable number of years. These cards are the only source of neonatal blood that is kept for all infants beyond the neonatal period.
We undertook this study to define an MR imaging pattern suggestive of congenital CMV infection and used PCR testing to detect CMV DNA on Guthrie cards for retrospective validation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MR Imaging Criteria
To define MR imaging criteria that should lead to a suspicion of congenital CMV infection, we analyzed the results of MR imaging examinations performed in the eight patients (five girls, three boys) in our institution who were known at that time to have proved congenital CMV infection. The diagnosis was established on the basis of isolation of the virus from urine, blood, or both within 3 weeks after birth. The test for CMV had been performed because of maternal primary CMV infection during pregnancy (two patients), signs of generalized CMV infection in the neonatal period (four patients), or neonatal signs of encephalopathy (two patients). If results from multiple MR imaging examinations were available, those from the most recent examination were used for scoring.
Unless myelination is substantially delayed, an assessment of white matter lesions is possible from the age of 8–10 months onward. After that age, T2-weighted MR images show decreased signal intensity of the white matter diffusely throughout the brain (21), against which background lesions stand out as bright. In all cases in which immature myelination prevented evaluation of the presence of white matter lesions, a follow-up MR imaging examination was performed. The mean age of the patients at the time of MR imaging was 22 months, with a range of 10–60 months (mean age and age range, respectively, 25 and 12–60 months for girls and 18 and 10–27 months for boys).
All MR images in this study were scored in consensus by two investigators (M.S.v.d.K. and F.B., with 16 and 12 years of experience with pediatric MR image interpretation, respectively), as described previously (1,4). Twenty parameters pertinent to the abnormalities observed in the patient population were scored (Table 1). On the basis of the findings in the eight patients with documented CMV infection, the provisional MR imaging criteria listed immediately below, which were the only criteria met by all eight patients, were established.
|
Patients, MR Imaging, and Guthrie Card Evaluation
Subsequently, the MR imaging criteria were applied as inclusion criteria to a group of 152 patients with clinically static encephalopathy of unknown origin; most of these patients had been referred because of white matter abnormalities of undetermined etiology. All MR images showed variable white matter abnormalities with or without cerebral gyral abnormalities. None of the 152 patients had shown signs of congenital CMV infection during the neonatal period. Many images had been sent to us for a second opinion from other centers, and, consequently, different pulse sequences had been used. However, at least T1-weighted sagittal and T2-weighted transverse MR images were available for all patients.
The MR imaging findings in 43 patients fulfilled the inclusion criteria. Twenty-one patients were excluded from further evaluation because their Guthrie cards had been destroyed. The MR images of the remaining 22 patients (11 girls, 11 boys) were evaluated by using the 20 parameters described in Table 1. At the time of their most recent MR imaging examination, the patients had ranged in age from 10 to 62 months, with a mean age of 24 months ± 12 (SD). The girls ranged in age from 10 to 43 months, with a mean age of 24 months ± 10; the boys ranged in age from 10 to 62 months, with a mean age of 23 months ± 14. In all patients, myelination was sufficient for the evaluation of white matter lesions on MR images. If results of multiple MR imaging examinations were available, results of previous MR imaging examinations were reviewed to determine the presence or absence of evidence of disease progression. The readers of the MR images were not aware of the results of the Guthrie card analysis.
In addition, 43 patients were randomly selected from the remaining 109 patients whose MR imaging results did not fulfill the inclusion criteria. Guthrie cards were available for 20 of these patients (seven girls, 13 boys). At the time of their most recent MR imaging examination, the patients had ranged in age from 12 to 52 months, with a mean age of 25 months ± 12. The girls ranged in age from 14 to 52 months, with a mean age of 26 months ± 16; the boys ranged in age from 12 to 45 months, with a mean age of 25 months ± 10.
Most patients came from the Netherlands; some patients came from other west European countries or North America. It is known that in Western Europe and North America, 0.5%–2% of neonates have congenital CMV infection (22,23). For this reason, we also analyzed the Guthrie cards of 300 neonates who were not suspected of having CMV infection and who were matched in terms of birth year with the patients but were otherwise randomly selected.
For each Guthrie card examined, we also analyzed the neighboring two cards—without identification of the donor children—because the possibility of contamination of Guthrie cards by neighboring CMV-positive cards has been suggested (18). For each card, a new disposable razor blade was used to cut out a full blood spot. Each blood spot was tested in duplicate. The PCR procedure was performed as described previously (24,25). If the internal control DNA signal was negative, the entire procedure was repeated for a new blood spot. The investigators involved in the Guthrie card analysis (G.V. performed the analysis with the supervision of J.F.L.W.) were not aware of the MR imaging results.
Statistical Analysis
Sex differences were evaluated by using the Fisher exact test; age differences were tested by using the Mann-Whitney test.
Univariate relationships between all MR imaging findings scored (except the findings constituting the inclusion criteria) and CMV status were examined for statistical significance by using the Fisher exact test for all 30 patients suspected of having or previously proved to have CMV infection. Of the 20 findings scored at MR imaging (Table 1), only those six (ie, number of lesions, consistency of lesion size, myelination, presence of gyral abnormalities, predominant location of gyral abnormalities, and presence of abnormalities of the anterior part of the temporal lobe) for which more than three patients had different scores from the modal category were used, so that a P value of less than .05 could be achieved at comparison of patients with positive PCR results for CMV infection and patients with negative PCR results with the Fisher exact test. Variables with more than two categories were dichotomized at cutoff values that maximized the number of patients per category.
Statistical analysis to evaluate the potential contribution of additional MR imaging criteria with respect to the diagnosis of congenital CMV infection was performed for the 22 patients who fulfilled the initial MR imaging criteria for congenital CMV infection and whose CMV status was determined with PCR analysis of the Guthrie card. To identify the optimal combination of variables predictive of PCR status, linear discriminant analysis was used in a backward elimination way. Starting with all six variables, variables were excluded sequentially until only one remained. The order in which the variables were eliminated was determined by the P value for PCR status found at an analysis of covariance where the specific variable being evaluated was included as the dependent variable and the other variables in the model were included as covariates.
The optimal number of variables was determined with the Youden index for a diagnostic tool (ie, the mean of sensitivity and specificity). The threshold used for the linear discriminant analysis score for predicting patients as having negative or positive PCR results was zero. For calculation of the Youden index, sensitivity and specificity were determined by using leave-one-out cross validation. That is, each patient was excluded in turn and his or her PCR status was predicted by using linear discriminant analysis on the basis of results in the remaining patients.
To obtain an unbiased estimate of the performance of the resulting linear discriminant analysis predictive score, a second loop of leave-one-out cross validation was performed. Again, each patient was excluded in turn, and the variables to be used for prediction of the PCR status of that patient were determined by using the procedure described above as applied to the remaining 21 patients. The resulting prediction from the linear discriminant analysis of the selected variables was then compared with the true PCR status of that patient. A combination of these comparisons across all 22 patients resulted in estimates of sensitivity and specificity. This procedure also enabled examination of the stability of the variable selection procedure described above.
CIs for proportions were calculated according to the method of Clopper and Pearson (26). P values were adjusted for multiple comparisons only where explicitly stated in the Results section. In those cases, the adjustment was performed by using the procedure of Hommel as adapted by Wright (27).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
Linear discriminant analysis of results for the remaining 21 patients who fulfilled the MR imaging criteria for possible congenital CMV infection resulted in the selection of abnormalities of the anterior part of the temporal lobe as the optimal predicting variable. This analysis did not reveal a combination of predictive variables that had better diagnostic power than the presence of abnormalities in the anterior part of the temporal lobe alone. Leave-one-out cross-validation analysis resulted in the correct prediction of PCR status on the basis of this finding for seven of 11 patients with a positive PCR result and for nine of 10 patients with a negative PCR result; this indicates that there was an association between prediction and true PCR status (P = .024, Fisher exact test).
When abnormalities of the anterior part of the temporal lobe were incorporated into the MR imaging inclusion criteria, the positive predictive value of these criteria was 89% (eight of nine patients; 95% CI: 52%, 99%), and the negative predictive value was 88% (29 of 33 patients; 95% CI: 72%, 97%). Overall, 15 of the 20 patients with proved congenital CMV infection (ie, the eight patients with neonatally documented CMV infection plus the 12 patients with positive PCR results) had abnormalities of the anterior part of the temporal lobe, while only one patient with a negative PCR result had such abnormalities.
In 11 of the 20 patients with proved congenital CMV infection, results of multiple MR imaging examinations were available. The mean interval between the first and the most recent MR imaging examination was 14 months (range, 7–23 months). In three children, the white matter lesions became better delineated as myelination progressed. Otherwise, the abnormalities remained unchanged in all patients, a finding that confirms the static nature of the encephalopathy.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The data described in the preceding paragraph indicate that congenital CMV infection is one of the leading causes of mental deficiency (22,23). The fact that congenital CMV infection frequently escapes detection in the neonatal period but may have neurologic sequelae underscores the need for a method to link intrauterine CMV infection with neurologic features such as unexplained leukoencephalopathy in children.
The introduction of PCR testing of blood stored on Guthrie cards as a standard of reference for the diagnosis of congenital CMV infection has created the opportunity for correct diagnosis beyond the neonatal period. Barbi et al (20) found in a large study population that the sensitivity of PCR testing for CMV DNA on Guthrie cards is 100% and the specificity is 99%. The only disadvantage of the test is that the PCR result is negative if the neonate no longer has viremia. The chance that a neonate with a congenital CMV infection no longer has viremia is very small but may exist, especially if the CMV infection occurred long before birth.
Most published reports of studies involving neuroimaging in congenital CMV infection concern computed tomographic (CT) scans obtained during follow-up of patients with neonatally symptomatic congenital CMV infection (29–31). Frequent findings are intracranial calcifications (in 33%–54% of patients), ventriculomegaly (in 10%–37% of patients), white matter abnormalities (in 0%–22% of patients), neuronal migration abnormalities (in 0%–10% of patients), and an extensive destructive encephalopathy (in 5%–13% of patients) (29–31). In 20%–30% of patients, no abnormalities are found (29,30). In patients with confirmed but neonatally asymptomatic congenital CMV infection, CT scans show milder abnormalities and show abnormalities less frequently. Williamson et al (32) observed white matter abnormalities in 14% of the children with asymptomatic congenital CMV infection in their study, whereas no abnormalities were demonstrated in 86% of the children. None of the children had calcifications or ventriculomegaly (32).
Studies of MR imaging findings during the follow-up of patients with neonatally symptomatic congenital CMV infection have been restricted to small numbers of patients (12–17). The findings in these patients have included dilated ventricles, enlarged subarachnoid spaces, gyral abnormalities, delayed myelination, and white matter lesions. To our knowledge, there has been no study of MR imaging findings in patients with proved but neonatally asymptomatic congenital CMV infection, but some researchers have reported MR imaging findings in patients with a presumed diagnosis (16,17). Barkovich and Lindan (16) reported the occurrence of cysts in the anterior portion of the temporal lobe and frequently observed dilated inferior horns. Although white matter abnormalities have been observed in congenital CMV infection, no particular pattern has been described (12–17).
We report a distinct pattern of MR imaging abnormalities in a group of patients with neonatally asymptomatic but proved congenital CMV infection. The pattern consists of white matter abnormalities either alone or in combination with gyral abnormalities in the form of polymicrogyria. In the absence of gyral abnormalities, the white matter abnormalities consist of multifocal lesions, with the largest lesions in the parietal area and predominantly involving the deep white matter, relatively sparing the immediately periventricular and subcortical white matter. In patients with gyral abnormalities, both diffuse and multifocal white matter abnormalities may occur. In addition, abnormalities of the anterior part of the temporal lobe, including abnormal and swollen white matter, cysts, and focal enlargement of the anterior part of the inferior horn—either alone or more often in combination—appear to be particularly suggestive of congenital CMV infection.
The MR imaging criteria we used were specifically designed to enable us to screen patients with leukoencephalopathy of unknown origin to determine if congenital CMV infection was a possible cause of their leukoencephalopathy. Screening tests in situations such as that described in this report should have a high negative predictive value, but a lower positive predictive value is acceptable. Strict initial MR imaging criteria would lead to the exclusion of patients with possible congenital CMV infection, who would not have their Guthrie cards analyzed. Our criteria worked well as an initial screening method. The fact that the PCR result was positive in as many as 55% of the patients who fulfilled the MR imaging criteria implies that we had possibly already missed patients with congenital CMV infection—most probably those with mild encephalopathy and no or only minimal white matter abnormalities. It is, however, reassuring that we did not find a positive PCR result in any patient with leukoencephalopathy of unknown origin who did not fulfill the MR imaging criteria.
It should be emphasized that the relatively low positive predictive value of the initial MR imaging criteria means that a confirmatory test is required for patients meeting the criteria. The PCR technique for revealing CMV DNA on Guthrie cards has sensitivity and specificity values that are close to 100% (20), making it ideal as a confirmatory test. The only major limitation of this approach is the fact that Guthrie cards are not kept indefinitely. When a definitive, PCR-based diagnosis is not possible, stricter MR imaging criteria with a higher positive predictive value (and not too great an effect on the negative predictive value) should be applied. Inclusion of the presence of abnormalities in the anterior portion of the temporal lobe may serve this purpose.
The most substantial limitation of the present study is the relatively small size of the patient population. Apart from the difference in the frequency of occurrence of abnormalities in the anterior portion of the temporal lobe, the difference in the frequency of occurrence of the MR imaging findings between the patients with positive PCR results and those with negative results did not reach the level of significance; this may be related to the limited statistical power with this relatively small sample. The predictive value of the other findings has to be confirmed in a larger study. A larger study will also result in smaller CIs for the positive and negative predictive values of the MR imaging criteria. In addition, the validity of the MR imaging criteria and the interobserver variability still have to be tested in a larger, independent series of patients with static leukoencephalopathy of unknown origin that includes patients who may have congenital CMV infection. Finally, our study was not designed to evaluate the occurrence and possible range of cerebral abnormalities in patients with congenital CMV infection in general. A prospective study in which a large cohort of neonates is tested for congenital CMV infection and followed up with MR imaging is needed to address this issue.
The MR imaging pattern suggestive of congenital CMV infection does not resemble the pattern of any of the known leukoencephalopathies (1,3,4), with the exception of the congenital muscular dystrophies (33). A combination of cortical dysgyria and diffuse or multifocal white matter abnormalities is also seen in muscle-eye-brain disease and in the Fukuyama type of congenital muscular dystrophy (33). However, in addition to muscle weakness, patients with the latter disease have other abnormalities at MR imaging, such as pontine hypoplasia and subcortical cerebellar cysts (33). None of these is seen in congenital CMV infection.
A combination of multifocal white matter abnormalities and cysts in the anterior portion of the temporal lobe has been reported previously in a few children with clinically static encephalopathy (34,35). It has been suggested that this may be a novel, genetically determined leukoencephalopathy. Given our findings, congenital CMV infection should also be considered.
In conclusion, MR imaging can be used to direct the diagnostic process in children with leukoencephalopathy. The MR imaging pattern suggestive of congenital CMV infection consists of white matter abnormalities either alone or in combination with gyral abnormalities in the form of polymicrogyria; in the absence of gyral abnormalities, the white matter abnormalities consist of multifocal lesions, with the largest lesions in the deep parietal area, relatively sparing the immediately periventricular and subcortical white matter; in the presence of gyral abnormalities, both diffuse and multifocal white matter abnormalities may occur. If this MR imaging pattern is present in a patient with a clinically static encephalopathy, we believe that PCR testing for CMV DNA on the Guthrie card should be the first diagnostic test. If the results of PCR testing are positive, random metabolic screening can be precluded. If the Guthrie card is not available, stricter MR imaging criteria, including the presence of abnormalities in the anterior part of the temporal lobe, should be applied to enable a probable MR imaging–based diagnosis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.S.v.d.K.; study concepts, M.S.v.d.K., A.A.M.H., J.F.L.W.; study design, M.S.v.d.K., J.F.L.W.; literature research, M.S.v.d.K., J.F.L.W.; clinical studies, M.S.v.d.K., F.B.; experimental studies, G.V., J.F.L.W., J.G.L.; data acquisition, M.S.v.d.K., G.V., F.B., J.G.L.; data analysis/interpretation, M.S.v.d.K., J.F.L.W., F.B.; statistical analysis, A.A.M.H.; manuscript preparation, M.S.v.d.K., F.B., A.A.M.H., J.F.L.W.; manuscript definition of intellectual content, M.S.v.d.K.; manuscript editing and final version approval, M.S.v.d.K., F.B.; manuscript revision/review, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. S. van der Knaap, W.F.M. Arts, J. Y. Garbern, G. Hedlund, F. Winkler, C. Barbosa, M. D. King, A. Bjornstad, N. Hussain, M. K. Beyer, et al. Cerebellar leukoencephalopathy: Most likely histiocytosis-related Neurology, October 21, 2008; 71(17): 1361 - 1367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Walter, C Atkinson, M Sharland, P Rice, E Raglan, V C Emery, and P D Griffiths Congenital cytomegalovirus: association between dried blood spot viral load and hearing loss Arch. Dis. Child. Fetal Neonatal Ed., July 1, 2008; 93(4): F280 - F285. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.S. van der Knaap, M. Kriek, W.C.G. Overweg-Plandsoen, K.B. Hansson, K. Madan, J.S. Starreveld, P. Schotman-Schram, F. Barkhof, and S.A.M.J. L. Oberstein Cerebral White Matter Abnormalities in 6p25 Deletion Syndrome AJNR Am. J. Neuroradiol., March 1, 2006; 27(3): 586 - 588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Blumkin, N. Watemberg, D. Lev, G. Malinger, Y. Luckman, B. Ben-Zeev, and T. Lerman-Sagie Nonprogressive Familial Leukoencephalopathy With Porencephalic Cyst and Focal Seizures J Child Neurol, February 1, 2006; 21(2): 145 - 148. [Abstract] [PDF]
|
|
|
|
|
|
B. Tatli, M. Ozmen, N. Aydinli, M. Caliskan, P. M. Wertheim-van Dillen, and M. S. van der Knaap Not a New Leukodystrophy but Congenital Cytomegalovirus Infection J Child Neurol, June 1, 2005; 20(6): 525 - 527. [Abstract] [PDF]
|
|
|
|
|
|
B. Tatli, M. Ozmen, N. Aydinli, M. Caliskan, P. M. Wertheim-van Dillen, and M. S. van der Knaap Not a New Leukodystrophy but Congenital Cytomegalovirus Infection J Child Neurol, June 1, 2005; 20(6): 525 - 527. [Abstract] [PDF]
|
|
|
|
|
|
M. Henneke, N. Preuss, V. Engelbrecht, F. Aksu, E. Bertini, G. Bibat, K. Brockmann, C. Hubner, M. Mayer, V. Mejaski-Bosnjak, et al. Cystic leukoencephalopathy without megalencephaly: A distinct disease entity in 15 children Neurology, April 26, 2005; 64(8): 1411 - 1416. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
