Prospective Identification of Congenital Cytomegalovirus Infection in Newborns Using Real-Time Polymerase Chain Reaction Assays in Dried Blood Spots

Transcription

1 MAJOR ARTICLE Prospective Identification of Congenital Cytomegalovirus Infection in Newborns Using Real-Time Polymerase Chain Reaction Assays in Dried Blood Spots Marianne Leruez-Ville, 1 Christelle Vauloup-Fellous, 4 Sophie Couderc, 5 Sophie Parat, 2 Christine Castel, 6 Véronique Avettand-Fenoel, 1 Tiffany Guilleminot, 1 Liliane Grangeot-Keros, 2 Yves Ville, 2 Sophie Grabar, 3 and Jean-Francxois Magny 7 1 Laboratoire de Virologie, Hôpital Necker-Enfants-Malades, Centre National de Référence du Cytomégalovirus-Laboratoire Associé, 2 Maternité.Hôpital Necker-Enfants-Malades, and 3 Unité de Biostatistique et Epidémiologie, Hôpital Cochin, Assistance Publique de Paris (APHP), Université Paris Descartes, Paris, France, 4 Service de Microbiologie-Immunologie biologique, Hôpital Antoine Béclère, Assistance Publique de Paris (APHP), Université Paris Sud, Clamart, France, 5 Maternité. Centre Hospitalier Intercommunal de Poissy-Saint Germain, Poissy, France, 6 Maternité. Hôpital Antoine Béclère, Assistance Publique de Paris (APHP), Clamart, France, and 7 Centre de Pédiatrie et de Réanimation Néonatale, Institut de Puériculture, Paris, France (See editorial commentary by Pass, pages ) Background. Congenital cytomegalovirus (CMV) infection is a public health issue, and implementation of neonatal screening has been debated. Detection of CMV DNA by polymerase chain reaction (PCR) of dried blood spots (DBS) routinely collected for metabolic screening from all newborns has been proposed for congenital CMV infection screening. The goal of this study was to prospectively assess the performance of 2 CMV PCR assays of DBS for CMV neonatal screening in a selected population of neonates. Methods. We studied prospective congenital CMV screening in a population of neonates either born with symptoms compatible with congenital CMV or born to mothers with a history of primary infection during pregnancy. For each neonate, 2 CMV PCR assays of DBS were blindly performed in parallel with a gold standard technique (ie, CMV PCR of a urine sample). Results. Two hundred seventy-one neonates were studied, and CMV infection, defined by a positive urine sample in the first week of life, was confirmed in 64 (23.6%). Nineteen infected (29.7%) neonates were symptomatic, and 45 (70.3%) were asymptomatic. The ranges of sensitivity, specificity, positive predictive value, and negative predictive value for the 2 CMV PCR assays of DBS were 95.0% 100%; 98.1% 99.0%; 94.1% 96.9%, and 98.5% 100%, respectively. Conclusions. The sensitivity and specificity of both CMV PCR assays of DBS to identify congenital CMV were very high in this population of neonates with a high risk of sequelae. These new data should be considered in the ongoing debate on the appropriateness of the use of DBS as a sample to screen for congenital CMV infection. Cytomegalovirus (CMV) is the leading cause of congenital infection in industrialized countries and may explain up to 15% 20% of sensorineural hearing loss in Received 1 July 2010; accepted 17 December Correspondence: Marianne Leruez-Ville, MD, Laboratoire de Virologie, Hôpital Necker-Enfants-Malades, Université Paris, Descartes, Paris, 75015, France (marianne.leruez@nck.aphp.fr). Clinical Infectious Diseases 2011;52(5): Ó The Author Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please journals.permissions@oup.com /2011/ $37.00 DOI: /cid/ciq241 infants and children [1, 2]. In the absence of an established treatment and of a consensual protocol for monitoring infected children, neither antenatal nor neonatal screening policies have been implemented in any industrialized country. However, both the high incidence of congenital CMV infection and its importance as a cause of sensorineural impairment are feeding the debate on the justification of neonatal screening for CMV infection. The 2 main potential benefits of neonatal screening are the treatment to prevent the onset or progression of sensorineural hearing loss and the PCR in DBS for Congenital CMV Screening d CID 2011:52 (1 March) d 575

2 identification of children at risk for late-onset or progressive sensorineural hearing loss [3]. CMV polymerase chain reaction (PCR) assays of dried blood spot (DBS) were first developed for the retrospective diagnosis of congenital infection in children with hearing loss or other congenital CMV- related symptoms that were not diagnosed at birth [4, 5, 6, 7, 8, 9, 10, 11] and have also been proposed for neonatal screening [12]. A number of protocols for CMV PCR of DBS have been reported in the literature [4, 5, 8, 9, 10, 13, 14, 15], with variable in vitro sensitivities depending on the DNA extraction method used [15]. In 2006, we validated 2 CMV PCR assays of DBS based on 2 different extraction and amplification protocols that demonstrated good in vitro sensitivity [13]. The objective of the present study was to prospectively assess the sensitivity, specificity, and positive and negative predictive values of 2 CMV PCR assays of DBS for identification of neonates with congenital CMV infection in a population of French neonates at high risk of infection, in comparison with standard urine detection. The results of the study are discussed in the light of the ongoing debate on using DBS as a sample to screen for congenital CMV infection. MATERIAL AND METHODS Study Population Prospective Study. Neonates were prospectively enrolled in 3 maternity units at the main referral centers for congenital CMV infection in the Paris area (Necker Hospital/Institut de Puériculture, Poissy Hospital, and Béclère Hospital) from January 2008 through December Participation in the study was offered to the parents of all neonates belonging to a high-risk group (ie, children with clinical symptoms compatible with congenital CMV and/or born to a mother with a history of primary CMV infection during pregnancy). Urine samples were collected during the first week of life. Guthrie cards were collected within 72 h of life for metabolic screening, stored in standard conditions, and retrieved from the regional screening laboratories after the completion of metabolic screening. Ethics committee approval was obtained (authorization A ). All parents gave written consent for their children to participate to the study. Retrospective Study in Children Born to HIV-Infected Mothers. Although neonatal screening for congenital CMV infection is not recommended in France, it is widely performed in children born to HIV-infected mothers [16]. When the result of CMV neonatal screening of a urine specimen was positive, parents were asked to participate in the study and Guthrie cards were retrieved from the regional screening laboratories. Methods of CMV Detection. The reference technique was a CMV PCR of a urine sample that was performed using either an in-house real-time PCR (Virology laboratories of Necker and Poissy hospitals), as elsewhere [17], or the CMV PCR kit (Abbott; Béclère Hospital). For DBS, CMV PCR assays were performed blinded to the results of the reference technique and systematically with 2 techniques (protocols 1 and 2) in 2 different laboratories (Necker and Béclère), as described elsewhere [13]. DNA Extraction from DBS. In brief, a whole single DBS (diameter, 10 mm) was cut in thin strips, which were used for DNA extraction. In protocol 1, after pretreatment with a lysis buffer (sodium hydroxide,.32 %), extraction was performed using the QIAamp DNA Blood Mini Kit (Qiagen), according to the manufacturer s instructions. DNA was eluted with 50 ll of elution buffer. In protocol 2, after incubation with a lysis buffer (Tris-HCl, 50 mm; ph, 7.5; sodium chloride, 50 mm; EDTA, 10 mm; SDS,.5%; proteinase K, 10 mg/ml), phenol-chloroform extraction and ethanol precipitation were performed. The DNA pellet was suspended in 25 ll of DNasefree buffer (Tris-HCl, 10 mm; EDTA, 1 mm). In protocol 1, all DNA extracts were submitted to an in-house human albumin gene real-time PCR assay [18]. This was done to check for the quality of DNA extraction from each sample. Moreover, the quantity of DNA extracted was measured using optic density at 260 nm (NanoDrop, Spectrophotometer, ND- 100, Labtech). CMV DNA Amplification. In protocol 1, amplification of CMV DNA was performed using an in-house real-time PCR, as described elsewhere [17], and all extracts were amplified in duplicate. The results obtained with this protocol were analyzed under 2 conditions: the neonate was considered to be infected when at least 1 of the 2 duplicates was positive (1/2 or 2/1; condition 1) or when the 2 duplicates tested positive (1/1; condition 2). In protocol 2, CMV DNA was amplified using the CMV PCR kit (Abbott) in a single amplification. Statistical Analysis. Sensitivity, specificity, and positive and negative predictive values were calculated using standard methods and exact 95% confidence limits. Agreement between the 2 protocols was evaluated using Cohen s Kappa coefficient. Comparisons between the 2 PCR assays were performed using the Mac Nemar test for paired data. CMV loads measured in neonatal DBS according to symptomatic and asymptomatic clinical status at birth were compared using the Wilcoxon rank sum test. All the statistical analysis were performed using SAS software, version 9.2 (SAS Institute). RESULTS Study Population Prospective Study. The description of the 271 neonates included in study is reported in Table 1. Reasons for neonatal screening were different in infected and noninfected neonates 576 d CID 2011:52 (1 March) d Leruez-Ville et al

3 (Fisher s exact test, P,.001). Most infected neonates (68.7%) were screened because of a history of maternal primary infection, and the most frequent reason for screening in the entire population was the presence of CMV-compatible symptoms (60.5%) (Table 1). Of the 64 infected neonates, 45 (70.3%) were asymptomatic and 19 (29.7%) were symptomatic. Symptoms included intrauterine growth retardation in 14 patients (the only symptom in 5 patients), neurological involvement in 7 patients, thrombocytopenia in 9 patients, hepatomegaly in 5 patients, ascitis in 1 patient, and enterocolitis in 1 patient. Two infected neonates were treated with ganciclovir at the time of neonatal screening. Urine samples and DBS were obtained for all 271 neonates. For 6 neonates, only 1 DBS was available after the completion of metabolic neonatal screening; for these 6 neonates, CMV PCR of DBS was performed only with protocol 1. Retrospective Study. During the study period, urine samples from 369 neonates born to HIV-infected mothers were screened for congenital CMV at Necker Virology Laboratory; 10 of these neonates had a positive detection in the urine (2.7%); none of these 10 neonates were HIV coinfected. For 7 of these 10 neonates, the Guthrie card was tested. Six of these neonates were asymptomatic; one had a severe infection with neurologic involvement. Sensitivity and Specificity of CMV PCR Assays of DBS (Table 2). With protocol 1, albumin PCR results were positive for all DBS extracts, and the mean (6 standard deviation) amount of genomic DNA was lg/ml. The performances of the 2 protocols are reported in Table 2. There was a good level of agreement between the 2 protocols (j,.90; 95% confidence interval,.84.96). With protocol 1, 2 DBS from infected infants were CMV PCR discordant in the 2 duplicates (1/2). These DBS were obtained from 2 asymptomatic neonates. With this protocol, 4 falsepositive results were observed (in 2 cases with both duplicate positive and in 2 cases with only 1 positive duplicate). With protocol 2, 3 false-negative results of CMV PCR of DBS were observed. Two of these 3 DBS were obtained from 2 symptomatic neonates presenting with isolated intrauterine growth retardation and 1 from an asymptomatic neonate. With this protocol, 3 false-positive results were observed. Results of CMV PCR of DBS from Neonates Born to HIV- Infected Mothers. All 7 DBS were positive with protocol 2, and at least 1 of the 2 duplicates was positive in all neonates with protocol 1; however, for 3 DBS, only 1 duplicate tested positive with this protocol. CMV DNA Quantification in DBS. With protocol 1, the median viral load in positive DBS was 3.87 log 10 copies/ml (range, log 10 copies/ml) (Table 2). Viral loads were also calculated as the number of CMV DNA copies per 150,000 cells, because albumin PCR allowed calculation of the number of cells studied in each PCR well. When expressed in this manner, the median viral load of positive DBS was 3.02 log 10 copies/ 150,000 cells (range, log 10 copies/150,000 cells). With protocol 2, the median viral load of positive DBS was 3.29 log 10 copies/ml (range, log 10 copies/ml). The correlation between viral loads obtained with the 2 protocols was weak (R 5.42). Viral loads in false-positive DBS were low with the 2 protocols. CMV DNA Loads in DBS and Neonatal Clinical Status (Figure 1). The viral loads obtained with both protocols were compared between symptomatic and asymptomatic neonates. For this analysis, only the patients whose DBS were tested with the 2 protocols were analyzed (n 5 60). Moreover, the viral load values obtained in the DBS from the 2 symptomatic neonates treated with intravenous ganciclovir at the time of screening were excluded from this analysis. Therefore, this analysis included 43 asymptomatic and 15 symptomatic neonates. With protocol 1, the median viral load was 3.76 log 10 copies/ ml (range, log 10 copies/ml) in asymptomatic neonates and 4.26 log 10 copies/ml (range, log 10 copies/ ml) in symptomatic neonates; the difference was statistically significant (Wilcoxon test, P 5.045). When viral loads were normalized with PCR albumin results, the median viral load in symptomatic neonates was also higher than that in asymptomatic neonates (3.00 log 10 copies/150,000 cells [range, log 10 copies/150,000 cells] vs 3.56 log 10 copies/150,000 cells [range, log 10 copies/150,000 cells]; P 5.052). On the contrary, with protocol 2, median viral loads were not significantly different (P 5.914), with 3.28 log 10 copies/ml (range, log 10 copies/ml) in asymptomatic neonates and 3.33 log 10 copies/ml (range, log 10 copies/ml) in symptomatic neonates. Analysis of Series from the Literature. In Table 3, we reported some characteristics of the 4 main published series comparing the performance of CMV PCR assays in DBS with that of a reference technique (CMV detection in urine or saliva samples). DISCUSSION The performance of 2 CMV PCR assays of DBS for diagnosis of congenital CMV infection was prospectively and blindly assessed in comparison with a reference technique (CMV PCR of a urine sample collected in the first week of life). The sensitivity and specificity of CMV PCR assays of DBS were very high (95% 96.9% and 98.5% 99%, respectively) and not statistically different between the 2 assays in the population screened. This population included neonates with symptoms compatible with congenital CMV infection and/or neonates born to mothers with a history of primary infection. In France and in some other European countries, CMV seroprevalence in pregnant women is PCR in DBS for Congenital CMV Screening d CID 2011:52 (1 March) d 577

4 Table 1. Characteristics of the 271 Infants Included in the Prospective Study Characteristic No. (%) of infants Sex Male 127 (46.9) Female 144 (53.1) Inclusion criteria Compatible CMV symptoms alone 164 (60.5) History of maternal primary infection 74 (27.3) Both previous reasons 19 (7.0) Not specified 14 (5.2) Infants with confirmed congenital 64 (23.6) CMV infection (positive reference technique) Clinical status at birth of infected infants Symptomatic 19 (29.7) Asymptomatic 45 (70.3) Inclusion criteria in infected neonates Compatible CMV symptoms alone 5 (7.8) History of maternal primary infection 44 (68.7) Both previous reasons 14 (21.9) Not specified 1 (1.6) NOTE. CMV, cytomegalovirus. 50% [19 22], and the risk of primary infection is estimated to be up to 1% of seronegative pregnant women [19, 21]. Although testing for CMV immune status during pregnancy is not recommended in France, a serologic test is often prescribed in the first trimester of pregnancy and pregnant women with primary infection are referred to specialized antenatal clinics for counseling and follow-up. The present study was performed in such clinical settings, explaining the inclusion of a large number of neonates born to mothers with a history of primary CMV infection. A number of studies have reported significantly higher levels of CMV DNA in whole blood in symptomatic newborns than in asymptomatic ones [23, 24, 25]. In the present study, viral loads in DBS from symptomatic neonates were significantly higher than those in asymptomatic neonates when viral loads were measured with the first protocol. We and others have previously shown that CMV DNA quantification in DBS based on QiaAmp DNA Blood Mini Kit extraction was highly correlated with DNA quantification in fresh whole blood [13, 26]. Therefore, some CMV PCR assays in DBS could potentially provide useful prognosis information, because a number of studies have reported the link between higher CMV load at birth and the presence of sequelae [13, 24, 26]. The main limitation of our study was that the evaluation of CMV PCR screening of DBS was not based on universal screening but on screening in a population of neonates at high risk of infection. Although sensitivity and specificity of our test should not be affected by differences in populations screened, the high positive predictive value obtained in our population at high risk cannot be transposed to the general population of neonates in which prevalence of congenital CMV infection is much lower (1%). Indeed, it is well known that the positive predictive value of a test is inevitably affected by the prevalence of the disease screened [27]. Moreover, in our study, 90% of infected neonates were born to mothers with a history of primary infection during pregnancy, and infected neonates born to mothers with secondary infection were probably underrepresented. No data are available on the respective proportions of infected French neonates who are born to mothers with primary infection and those who are born to mothers with secondary infection. However, in a very similar European population (with a seroprevalence in pregnant women of 50%), Table 2. Performance of Real-Time Polymerase Chain Reaction (PCR) Assays of Dried Blot Spots (DBS) to Detect Confirmed Congenital Cytomegalovirus (CMV) Infection in Neonates and Viral Loads in DBS Protocol 1 Protocol 2 At least 1 positive duplicate (1/2 or 2/1) Two positive duplicate (1/1) Reference test result Positive Negative Total Positive Negative Total Positive Negative Total Positive Negative Analyses, % (95% confidence interval) Sensitivity 100 ( ) 96.9 ( ) 95.0 ( ) Specificity 98.1 ( ) 99.0 ( ) 98.5 ( ) PPV 94.1 ( ) 96.9 ( ) 95.0 ( ) NPV 100 ( ) 99.0 ( ) 98.5 ( ) Viral load, median, log 10 copies/ml Confirmed positive (n5 64) 3.87 ( ) (n560) 3.29 ( ) False-positive (n54) 2.55 ( ) (n53) 1.38 ( ) At least 1 positive duplicate (1/2 or 2/1) (n52) 1.91 ( ) NA NOTE. NA, not applicable; NPV, negative predictive value; PPV, positive predictive value. 578 d CID 2011:52 (1 March) d Leruez-Ville et al

5 Figure 1. Cytomegalovirus (CMV) DNA loads in dried blood spots (DBS), according to neonatal clinical status. Results were obtained with protocol 1 and are expressed as log 10 copies per ml of blood. The boxes represent the interquartile range (IQR), with horizontal bars for median values and full circle for mean values. Vertical bars represent 1.5 IQR; the upper horizontal bar is the maximum value within 1.5 (IQR) above the 75th percentile, and lower horizontal bar is the minimum value within 1.5 (IQR) below the 25th percentile. Open squares are the maximum values above 1.5 (IQR) of the 75th percentile or the minimum value below 1.5 (IQR) of the 25th percentile..80% of neonates with infection detected by universal screening of a urine sample at birth were born to mothers with a history of primary infection during pregnancy; therefore, neonatal infection after secondary maternal infection accounted for,20% of all cases of congenital CMV infection [21]. This situation is different in countries with intermediate or high seroprevalence among pregnant women, where nonprimary infections account for a much larger proportion of congenital infections [28, 29]. To evaluate the performance of our CMV PCR assays of DBS in a population of neonates born to mothers from a highly immune population (.90% seroprevalence), we retrospectively studied neonatal cards of neonates born to mothers with HIV infection and screened CMV positive by PCR of a urine sample. During the study period, the prevalence of congenital CMV infection in this population was 2.7% (10 of 369 patients), in agreement with data previously published by our group [16]. In addition, in this small series, all DBS tested positive with the 2 CMV PCR assays. However, although this population of CMV-infected neonates born to HIV-infected mothers was not HIV coinfected, it might not be representative of the general population of CMV-infected neonates born to women with secondary CMV infection during pregnancy. Indeed, the HIV-positive status of the mothers increases the risk of congenital CMV infection [16] and might also influence neonatal viral load. Sensitivity of CMV PCR assays of DBS for the diagnosis of congenital CMV infection that was reported in the 4 largest series published in the literature ranged from 34% to 100% [4, 14, 26, 30]. Because comparison among studies is problematic [31], to better understand the wide range of sensitivity reported, we compared the main characteristics of these series (Table 3). This comparison allowed identification of 3 main reasons for the discrepancies among series: the size of DBS used (a larger size could account for higher sensitivity), the differences in PCR assays used, and the differences in populations tested. The characteristics of the population screened seem to influence the sensitivity of diagnosis of congenital CMV infection in DBS. In populations of neonates with symptoms or born to mothers who had a confirmed or suspected primary CMV infection during pregnancy, the sensitivity of CMV PCR of DBS was very high (Table 2). The sensitivity of CMV PCR of the DBS obtained in congenitally infected newborns detected through a universal screening program [21] was also as high as 82% in a study realized in a country where pregnant woman have a relatively low CMV seroprevalence [14]. Of interest, in this study, a better sensitivity of diagnosis of congenital CMV infection in DBS from neonates with congenital infection originating from maternal primary infection during pregnancy than in neonates infected after maternal recurrent infections (82% vs 54%) was reported. This difference may be explained by the fact that neonates born to mothers with recurrent infections may have a low viral load that could remain undetected by PCR of DBS. In the recent and important study from Boppana et al [30], PCR of DBS had a poor sensitivity (34%) in a population of neonates with infection detected through CMV universal screening of 20,448 newborns. The proportion of neonates infected after secondary maternal infections was not reported in this study but was likely to be high, because neonates in this study were born to mothers belonging in majority to ethnic groups with high CMV seroprevalence [32]. However, the major factor for discrepancies between studies is probably the differences in DNA extraction methods from DBS [33]. Indeed, evaluations of different protocols of CMV DNA extraction and amplification from DBS have clearly shown that some methods are more sensitive than others; comparison of 8 different DNA extraction methods reported sensitivities ranging from 32% to 73% [15]. Different approaches have been proposed for congenital CMV infection screening: detection of all infected newborns [4, 30] and detection of infected neonates who failed newborn hearing screening [34]. Here, we proposed a screening strategy that may not detect all infected newborns but that should detect infected neonates at high risk of sequelae. Indeed, although the best methods for DBS screening would miss cases of infected neonates with low or no viremia, it was shown that most neonates who will further develop sequelae are viremic at birth [24, 25, 26, 35]; long-term sequelae having been reported only occasionally in neonates with low-level or no viremia [36]. Therefore, CMV PCR of DBS is likely to be a very efficient method to identify neonates at high risk of long-term sequelae. If implementation of large-scale newborn screening for CMV is to be considered, PCR in DBS for Congenital CMV Screening d CID 2011:52 (1 March) d 579

6 Table 3. Major Series Published Comparing Cytomegalovirus (CMV) Polymerase Chain Reaction (PCR) Assays of Dried Blood Spots (DBS) with a Reference Technique for Diagnosis of Neonatal CMV Infection Size of DBS used, (total diameter) Type of extraction Sensitivity, % Specificity, % Neonates with confirmed CMV infection Seroprevalence in pregnant women in the country or ethnic group, % Inclusion criteria Total population screened Study Design Heat shock disks (9 mm) Retrospective [37] -Compatible symptoms -Maternal PI 83 NA Phenol chloroform Retrospective UN 50 [21] Universal screening 55 Whole spot (10 mm) Barbi et al (2000) [4] Soetens et al (2008) [14] 73 NA Easy MAG (BioMérieux) NA QiaAmp DNA Blood mini kit (Qiagen) Retrospective NA UN Confirmed CMV infection 70 Half spot (5 mm) Atkinson et al (2009) [38] Qiagen M48 (Qiagen) Prospective 20, [32] Universal screening 92 2 disks (6 mm) Boppana et al (2010) [30] Whole spot (10 mm) Present study Prospective [22] -Compatible symptoms -Maternal PI Modified QiaAmp DNA Blood mini kit (Qiagen) Phenol chlorofom NOTE. NA, not applicable; PI, primary infection; UN, unknown. the use of DBS samples would be the most practical way and it could easily be implemented together with the current universal newborn screening program [32]. On the basis of the encouraging results reported here, further studies are needed to evaluate the sensitivity of high-throughput screening by CMV PCR of DBS to identify the neonates that will later develop sequelae. Acknowledgments We thank Pr Paul Czernichow and Mrs Valérie Gauthreau from the Association Francxaise pour le Dépistage et la Prévention des Handicaps de l Enfant, for their help to retrieve Guthrie cards from the regional laboratories, and the parents of the neonates for their agreement to participate to the study. Financial support. This work was supported by La Fondation de l Avenir (Program: ET6-430). Potential conflicts of interest. LG-K has received payment for lectures from DiaSorin. CV-F has received grant support from BioMérieux. YV s institution has received grant support in his name from Direction de la Recherche Clinique and is a consultant for Sequenom. All other authors: no conflicts. References 1. Barbi M, Binda S, Caroppo S, Ambrosetti U, Corbetta C, Sergi P. A wider role for congenital cytomegalovirus infection in sensorineural hearing loss. Pediatr Infect Dis J 2003; 22: Grosse SD, Ross DS, Dollard SC. Congenital cytomegalovirus (CMV) infection as a cause of permanent bilateral hearing loss: a quantitative assessment. J Clin Virol 2008; 41: Grosse SD, Dollard S, Ross DS, Cannon M. Newborn screening for congenital cytomegalovirus: Options for hospital-based and public health programs. J Clin Virol 2009; 46(Suppl 4):S Barbi M, Binda S, Primache V, et al. Cytomegalovirus DNA detection in Guthrie cards: a powerful tool for diagnosing congenital infection. J Clin Virol 2000; 17: Distefano AL, Alonso A, Martin F, Pardon F. Human cytomegalovirus: detection of congenital and perinatal infection in Argentina. BMC Pediatr 2004; 4: Fischler B, Rodensjo P, Nemeth A, Forsgren M, Lewensohn-Fuchs I. Cytomegalovirus DNA detection on Guthrie cards in patients with neonatal cholestasis. Arch Dis Child Fetal Neonatal Ed 1999; 80:F Haginoya K, Ohura T, Kon K, et al. Abnormal white matter lesions with sensorineural hearing loss caused by congenital cytomegalovirus infection: retrospective diagnosis by PCR using Guthrie cards. Brain Dev 2002; 24: Johansson PJ, Jonsson M, Ahlfors K, Ivarsson SA, Svanberg L, Guthenberg C. Retrospective diagnostics of congenital cytomegalovirus infection performed by polymerase chain reaction in blood stored on filter paper. Scand J Infect Dis 1997; 29: Scanga L, Chaing S, Powell C, et al. Diagnosis of human congenital cytomegalovirus infection by amplification of viral DNA from dried blood spots on perinatal cards. J Mol Diagn 2006; 8: Yamamoto AY, Mussi-Pinhata MM, Pinto PC, Figueiredo LT, Jorge SM. Usefulness of blood and urine samples collected on filter paper in detecting cytomegalovirus by the polymerase chain reaction technique. J Virol Methods 2001; 97: Zucca C, Binda S, Borgatti R, et al. Retrospective diagnosis of congenital cytomegalovirus infection and cortical maldevelopment. Neurology 2003; 61: Barbi M, Binda S, Caroppo S. Diagnosis of congenital CMV infection via dried blood spots. Rev Med Virol 2006; 16: Vauloup-Fellous C, Ducroux A, Couloigner V, et al. Evaluation of cytomegalovirus (CMV) DNA quantification in dried blood spots: 580 d CID 2011:52 (1 March) d Leruez-Ville et al

7 retrospective study of CMV congenital infection. J Clin Microbiol 2007; 45: Soetens O, Vauloup-Fellous C, Foulon I, et al. Evaluation of different cytomegalovirus (CMV) DNA PCR protocols for analysis of dried blood spots from consecutive cases of neonates with congenital CMV infections. J Clin Microbiol 2008; 46: de Vries JJ, Claas EC, Kroes AC, Vossen AC. Evaluation of DNA extraction methods for dried blood spots in the diagnosis of congenital cytomegalovirus infection. J Clin Virol 2009; 46(Suppl 4):S Guibert G, Warszawski J, Le Chenadec J, et al. Decreased risk of congenital cytomegalovirus infection in children born to HIV-1-infected mothers in the era of highly active antiretroviral therapy. Clin Infect Dis 2009; 48: Leruez-Ville M, Ouachee M, Delarue R, et al. Monitoring cytomegalovirus infection in adult and pediatric bone marrow transplant recipients by a real-time PCR assay performed with blood plasma. J Clin Microbiol 2003; 41: Laurendeau I, Bahuau M, Vodovar N, et al. TaqMan PCR-based gene dosage assay for predictive testing in individuals from a cancer family with INK4 locus haploinsufficiency. Clin Chem 1999; 45: Gratacap-Cavallier B, Bosson JL, Morand P, et al. Cytomegalovirus seroprevalence in French pregnant women: parity and place of birth as major predictive factors. Eur J Epidemiol 1998; 14: Gaytant MA, Galama JM, Semmekrot BA, et al. The incidence of congenital cytomegalovirus infections in The Netherlands. J Med Virol 2005; 76: Naessens A, Casteels A, Decatte L, Foulon W. A serologic strategy for detecting neonates at risk for congenital cytomegalovirus infection. J Pediatr 2005; 146: Vauloup-Fellous C, Picone O, Cordier AG, et al. Does hygiene counseling have an impact on the rate of CMV primary infection during pregnancy? Results of a 3-year prospective study in a French hospital. J Clin Virol 2009; 46(Suppl 4);S Revello MG, Zavattoni M, Baldanti F, Sarasini A, Paolucci S, Gerna G. Diagnostic and prognostic value of human cytomegalovirus load and IgM antibody in blood of congenitally infected newborns. J Clin Virol 1999; 14: Boppana SB, Fowler KB, Pass RF, et al. Congenital cytomegalovirus infection: association between virus burden in infancy and hearing loss. J Pediatr 2005; 146: Lanari M, Lazzarotto T, Venturi V, et al. Neonatal cytomegalovirus blood load and risk of sequelae in symptomatic and asymptomatic congenitally infected newborns. Pediatrics 2006; 117:e Walter S, Atkinson C, Sharland M, et al. Congenital cytomegalovirus: association between dried blood spot viral load and hearing loss. Arch Dis Child Fetal Neonatal Ed 2008; 93:F Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ 1994; 309: Boppana SB, Fowler KB, Britt WJ, Stagno S, Pass RF. Symptomatic congenital cytomegalovirus infection in infants born to mothers with preexisting immunity to cytomegalovirus. Pediatrics 1999; 104: Mussi-Pinhata MM, Yamamoto AY, Moura Brito RM, et al. Birth prevalence and natural history of congenital cytomegalovirus infection in a highly seroimmune population. Clin Infect Dis 2009; 49: Boppana SB, Ross SA, Novak Z, et al. Dried blood spot real-time polymerase chain reaction assays to screen newborns for congenital cytomegalovirus infection. JAMA 2010; 303: Dollard S, Schleiss MR, Grosse SD. Public health and laboratory considerations regarding newborn screening for congenital cytomegalovirus. J Inherit Metab Dis 2010; 33: Kharrazi M, Hyde T, Young S, Amin MM, Cannon MJ, Dollard SC. Use of screening dried blood spots for estimation of prevalence, risk factors, and birth outcomes of congenital cytomegalovirus infection. J Pediatr 2010; 157: Dollard SC, Schleiss MR. Screening newborns for congenital cytomegalovirus infection. JAMA 2010; 304:407 8; author reply Choi KY, Schimmenti LA, Jurek AM, et al. Detection of cytomegalovirus DNA in dried blood spots of Minnesota infants who do not pass newborn hearing screening. Pediatr Infect Dis J 2009; 28: Ross SA, Novak Z, Fowler KB, Arora N, Britt WJ, Boppana SB. Cytomegalovirus blood viral load and hearing loss in young children with congenital infection. Pediatr Infect Dis J 2009; 28: Bradford RD, Cloud G, Lakeman AD, et al. Detection of cytomegalovirus (CMV) DNA by polymerase chain reaction is associated with hearing loss in newborns with symptomatic congenital CMV infection involving the central nervous system. J Infect Dis 2005; 191: Barbi M, Binda S, Caroppo S, et al. Multicity Italian study of congenital cytomegalovirus infection. Pediatr Infect Dis J 2006; 25: Atkinson C, Walter S, Sharland M, et al. Use of stored dried blood spots for retrospective diagnosis of congenital CMV. J Med Virol 2009; 81: PCR in DBS for Congenital CMV Screening d CID 2011:52 (1 March) d 581