Multicenter Evaluation of the DiaSorin Molecular Simplexa Congenital CMV Direct PCR Test on Neonatal Saliva and Urine Specimens

ABSTRACT

Cytomegalovirus (CMV) is the most common virus associated with congenital infection worldwide and is a major cause of sensorineural hearing loss (SNHL) and developmental delay. Up to 90% of infants with congenital CMV (cCMV) infection are asymptomatic at birth, making the diagnosis challenging. Postnatal diagnosis involves testing newborn saliva and/or urine collected before 21 days of life to confirm cCMV infection. This multicenter study evaluated the performance of the Simplexa Congenital CMV Direct real-time PCR assay for the qualitative detection of CMV in newborn saliva (n = 2,023) and urine (n = 1,797) specimens. Compared to two PCR/bidirectional sequencing assays, the Simplexa Congenital CMV Direct assay demonstrated positive percent agreement (PPA) and negative percent agreement (NPA) of 98.6% and 99.9%, respectively, for saliva samples and a PPA of 97.8% and an NPA of 99.9% for urine specimens. Overall concordance was κ = 0.98 or near perfect compared to the composite reference methods with both sample types. By 95% probit analysis, the limit of detection (LoD) using the AD-169 reference strain was 350 ± 12 copies/mL in urine. The LoDs of saliva swabs in either 1 mL or 3 mL of transport medium were 274 ± 12 copies/mL and 300 ± 14 copies/mL, respectively. The Simplexa Congenital CMV Direct assay can be applied to both saliva and urine specimens collected from newborns less than 21 days of age to rapidly and reliably identify CMV infection.

INTRODUCTION

Congenital cytomegalovirus (cCMV) infection is the most common intrauterine infection in the United States. The incidence of cCMV is approximately 5 to 7 per 1,000 live births in developed countries and is significantly higher in developing parts of the world where population seroprevalence is generally higher (1). Primary infection in the seronegative mother is uncommon but is associated with the greatest risk of transplacental transmission, and symptomatic infection in newborns occurs most frequently in this scenario (2). However, most transplacental transmission events are thought to result from nonprimary maternal infections and occur in the third trimester (3). The risk of long-term sequelae to the fetus in this situation is low. Transmission in the first trimester occurs less frequently but is typically associated with a higher rate of long-term sequelae in the newborn (4, 5).

Most infants with cCMV are asymptomatic during the newborn period, but about 10% to 15% of infected asymptomatic babies will go on to develop sensorineural hearing loss (SNHL). In symptomatic newborns, the mortality rate is 7% to 12%, and most suffer from permanent sequelae, such as SNHL, neurodevelopmental delays, retinitis, or cerebral palsy (6). The most common signs of symptomatic cCMV in neonates are jaundice, petechiae, hepatosplenomegaly, and microcephaly (7, 8). Brain imaging is critical to assess for findings of periventricular calcifications, ventricular dilatation, and cysts (9). These newborns usually have elevated aspartate aminotransferase, hyperbilirubinemia, thrombocytopenia, and elevated cerebrospinal fluid protein (8).

Postnatal diagnosis of cCMV involves testing newborn saliva, urine, or blood specimens collected before 21 days of age. Beyond about 3 weeks of age, it is difficult to determine if the infant could have contracted the infection through nursing or by exposure to other individuals shedding the virus. Early detection of infection and initiation of antiviral therapy can improve hearing and developmental outcomes. However, most states in the United States do not have universal newborn screening for cCMV; some require targeted testing only for those infants who fail their newborn hearing screen (NHS), or infants may be tested if they have signs and symptoms compatible with cCMV infection. However, because up to 90% of infants with cCMV are asymptomatic during the newborn period, many of these infections are not recognized until later in life.

Several molecular tests for CMV are FDA cleared/approved for determining viral load in plasma specimens, and laboratory-developed tests (LDTs) have been used to detect CMV in various sample types. Only one test is FDA cleared and commercially available for testing newborn saliva swab samples (10). Unfortunately, the positive predictive value of saliva testing for detection of actual cCMV infection is low due to the potential for contamination from breastmilk or maternal genital secretions, and an initial positive result requires confirmation using a urine specimen (11–13). In this study, we performed a clinical evaluation of the Simplexa Congenital CMV Direct real-time PCR assay to detect CMV in both saliva swabs and urine specimens obtained from infants younger than 21 days of age.

MATERIALS AND METHODS

Specimens.

Both prospective and retrospective specimens were included in the study. Clinical samples included saliva swabs and urine specimens collected from infants less than 21 days of age. Among 13 U.S. and two non-U.S. sites, 1,859 saliva swabs and 1,656 urine specimens were prospectively collected and were tested fresh or frozen (Table S1 in the supplemental material, prospective fresh and prospective frozen categories). Retrospective positive and negative frozen saliva swabs and urine specimens were obtained from one collection site (Università di Bologna) and included 53 CMV-positive and 120 CMV-negative saliva swab specimens as well as 53 CMV-positive and 120 CMV-negative urine specimens. Each site obtained institutional review board (IRB) approval or waiver based on its role in the study. All residual samples were deidentified before enrollment in the study and testing. It was unknown if infants had symptoms compatible with cCMV at the time of specimen enrollment.

Specimen collection and storage.

Nylon-flocked swabs were used to collect saliva specimens. All newborn saliva swabs were collected more than 1 h after breastfeeding. These saliva swabs were transported in 1 mL or 3 mL of transport medium, which consisted of one of the following depending on the enrollment site: BD universal viral transport (UVT), BD Liquid Amies (ESwab; BD, Franklin Lakes, NJ), Copan universal transport medium (UTM; Copan, Murietta, CA), Remel M4, Remel M4RT, or Remel M6 (Remel, Lenexa, KS). Urine specimens were collected without preservatives and either tested or aliquoted and frozen as follows: within 2 h after collection if stored at 18 to 25°C, within 8 h after collection if transported on ice or stored at 2 to 8°C, more than 8 h after collection if stored for up to 3 days at −20 ± 10°C, and more than 3 days after collection if stored at ≤−70°C. Saliva swab specimens were tested or aliquoted and frozen within 48 h after collection if stored at 18 to 25°C or within 7 days after collection if transported on ice and stored at 2 to 8°C. If there was a delay of more than 7 days before aliquoting the specimen, the specimens were kept for up to 14 days at −20 ± 10°C. If there was a delay of more than 14 days before testing, samples were stored at ≤−70°C.

For fresh specimens enrolled in the study (prospective fresh), three aliquots were prepared for each specimen: one aliquot for Simplexa testing, one aliquot for composite reference method (CRM) testing by sequencing, and one aliquot that was retained. Prospective fresh specimens were tested by Simplexa without freezing according to the storage conditions listed above. For prospective frozen and retrospective categories, specimens were thawed, and three aliquots were prepared for each specimen as described above before refreezing and shipment to testing sites.

DiaSorin Molecular Simplexa Congenital CMV Direct assay.

The Simplexa Congenital CMV Direct assay is a real-time PCR system that enables the amplification and qualitative detection of CMV DNA directly from saliva swabs in viral transport medium (VTM) and raw urine collected from neonates less than 21 days old without a separate nucleic acid extraction step. The PCR targets a highly conserved region of the CMV UL83 gene, and each reaction includes an internal control to assess amplification failure and/or potential inhibition. Briefly, 50 μL of the Simplexa Congenital CMV Direct reaction mix is added to the reaction (R) well in one wedge of the direct amplification disc (DAD), followed by the addition of 50 μL of the specimen (saliva swab VTM or urine) directly to the sample well. The wedge is sealed, and the DAD is loaded onto the Liaison MDX instrument for real-time PCR amplification and detection. After completion of the run, the software automatically calculates and displays the results. A detected result indicates the presence of CMV DNA in the patient sample, whereas no detectable result indicates the absence of CMV DNA above the limit of detection.

Comparison methods.

For prospective saliva swabs and urine, one central laboratory performed both comparator PCR/bidirectional sequencing assays for the two-part composite reference method (CRM). The CRM used two validated PCRs, followed by bidirectional sequencing assays. PCR/bidirectional sequencing assay 1 was designed to target the CMV UL54 gene, while PCR/bidirectional sequencing assay 2 was designed to target a different region of CMV UL83 than the region targeted by the Simplexa Congenital CMV Direct assay. The sequencing assays were performed in the following four steps: (i) internal control DNA template was added to the specimen, (ii) sample DNA was extracted using either a MagNA Pure (Roche, Indianapolis, IN) or Monarch extraction kit (New England Biolabs, Ipswich, MA), (iii) extracted samples were amplified using a QuantiTect SYBR green PCR kit (Qiagen) with primers specific for either UL54 or UL83, and (iv) a separate PCR for the internal control was set up to monitor PCR inhibition and to validate negative CMV results. Samples that were detected and generated a PCR amplicon with a defined melting temperature proceeded into sequencing. The PCR amplicons were bidirectionally sequenced using BigDye cycle sequencing chemistry (Life Technologies). DNA alignments of bidirectional sequencing results were compared to the GenBank database using a BLAST search to determine if the sample was positive for CMV. For a sample to be identified as CMV positive by the CRM, at least one PCR/bidirectional sequencing assay was required to be positive. A sample was deemed CMV negative if both PCR/bidirectional sequencing assays were negative.

Analytical sensitivity.

The limit of detection (LoD) was determined using quantified stocks of three CMV strains (AD169, Towne, and Merlin; Zeptometrix, Buffalo, NY) serially diluted in CMV-negative human saliva swabs in UTM with 1 mL of transport medium, UTM with 3 mL of transport medium, and M4RT with 3 mL of transport medium and urine matrices. The LoD was determined by two methods: positive rate and probit analyses. The positive rate was determined as the lowest dilution at which all replicates resulted in CMV positive with a greater than 95% detection rate. The LoD by probit was determined as the lowest detectable dilution at which the quantified CMV strains (copies/mL) resulted positive with a 95% probability of detection.

Statistical methods.

Percent positive and negative agreement, kappa, probit, and two-sided (upper/lower) 95% confidence interval (CI) were calculated using Microsoft Office Excel 365 MSO software (Microsoft, Redmond, WA). The positive percent agreement (PPA) was calculated as TP/(TP + FN) × 100, and the negative percent agreement (NPA) was calculated as TN/(TN + FP) × 100, where TP represents true-positive results, FN represents false-negative results, TN represents true-negative results, and FP represents false-positive results. Cohen’s kappa values (κ) were also calculated as a measure of the overall agreement, with values categorized as almost perfect (>0.90), strong (0.80 to 0.90), moderate (0.60 to 0.79), weak (0.40 to 0.59), minimal (0.21 to 0.39), or none (0 to 0.20) (14, 15). Probit analyses were used for the copies/mL determination of the analytical sensitivity study. The dose-response 95th percentile (with 95% CI) model was assessed using the Finney and Stevens calculations (16).

RESULTS

Saliva clinical testing.

A total of 1,853 prospective saliva swab specimens collected from infants less than 21 days of age were assessed using the Simplexa Congenital CMV Direct assay. The results were compared to the CRM, which consisted of two PCR/bidirectional sequencing assays performed on the same samples. The Simplexa Congenital CMV Direct assay detected CMV in 16 of 17 samples for a PPA of 94.1% (95% CI of 0.71 to 1.00) (Table 1). The one false-negative saliva swab sample was also negative in one of the two sequencing assays. The Simplexa Congenital CMV Direct assay detected CMV with a cycle threshold (C_T) of 37.0 in one additional sample that was negative by the CRM (NPA = 99.9% [95% CI of 0.99 to 1.00]). The PPA and NPA were 100% for 53 CMV-positive and 117 CMV-negative retrospective saliva swab specimens (Table 1). Overall, for both prospective and retrospective samples, the PPA and NPA were 98.6% (95% CI of 0.92 to 1.00) and 99.9% (95% CI of 0.99 to 1.00), respectively, with a near-perfect concordance and κ of 0.98 (95% CI of 0.97 to 1.00).

TABLE 1.

Saliva swab prospective and retrospective results: Simplexa Congenital CMV Direct versus CRM (n = 2,023)

	CRMa		(±95% CI)b,c
Simplexa Congenital CMV Direct	Positive	Negative	PPA	NPA	κd
Prospective samples
Positive	16	1e	94.1% (0.73 to 0.99)	99.9% (0.99 to 1.00)	0.94 (0.89 to 1.00)
Negative	1f	1,835
Retrospective samples
Positive	53	0	100% (0.93 to 1.00)	100.0% (0.97 to 1.00)	1.0 (0.99 to 1.00)
Negative	0	117
Total
Positive	69	1	98.6% (0.92 to 1.00)	99.9% (0.99 to 1.00)	0.98 (0.97 to 1.00)
Negative	1	1,952

^aComposite reference method (CRM) was based on the results from two PCR/bidirectional sequencing assays. A true positive was positive by at least one bidirectional sequencing assay. A true negative was determined if both bidirectional sequencing assays were negative.

^b±, upper/lower 95%.

^cCI, confidence interval; PPA, positive percent agreement; NPA, negative percent agreement.

^dAlmost perfect (>0.90), strong (0.80 to 0.90), moderate (0.60 to 0.79), weak (0.40 to 0.59), minimal (0.21 to 0.39), or none (0 to 0.20).

^eOne saliva swab sample was negative by the CRM. The Simplexa Congenital CMV Direct assay had a cycle threshold (C_T) of 37.0.

^fOne saliva swab sample was negative by CMV sequencing assay 1 and positive by CMV sequencing assay 2. The routine result from the study site lab-developed PCR was negative for CMV.

Urine clinical testing.

The Simplexa Congenital CMV Direct assay was used to assess 1,624 prospectively collected urine samples from infants less than 21 days of age. Compared to the CRM, the Simplexa assay detected CMV in 41 of 43 positive specimens (PPA = 95.3% [95% CI of 0.85 to 0.99]) and was negative for 1,581 specimens (NPA = 100% [95% CI of 0.99 to 1.00]) (Table 2). The two Simplexa assay false negatives were also negative in one of the two sequencing assays. For the 49 CMV-positive and 124 CMV-negative retrospective urine specimens, the Simplexa PPA was 100% (95% CI of 0.93 to 1.00), and the NPA was 98.4% (95% CI of 0.94 to 1.00). Two retrospective urine samples were negative by both PCR/bidirectional sequencing assays but CMV positive by the Simplexa Congenital CMV Direct assay, with C_T values of 34.8 and 36.8. For urine samples overall, the PPA and NPA were 97.8% (95% CI of 0.92 to 1.00) and 99.9% (95% CI of 0.99 to 1.00), respectively, with a near-perfect concordance and κ of 0.98 (95% CI of 0.96 to 1.00) (Table 2).

TABLE 2.

Urine prospective and retrospective results: Simplexa Congenital CMV Direct versus CRM (n = 1,797)

	CRMa		(±95% CI)b,c
Simplexa Congenital CMV Direct	Positive	Negative	PPA	NPA	κd
Prospective samples
Positive	41	0	95.3% (0.85 to 0.99)	100% (0.99 to 1.00)	0.98 (0.94 to 1.00)
Negative	2e	1,581
Retrospective samples
Positive	49	2f	100% (0.93 to 1.00)	98.4% (0.94 to 1.00)	0.97 (0.93 to 1.00)
Negative	0	122
Total
Positive	90	2	97.8% (0.92 to 1.00)	99.9% (0.99 to 1.00)	0.98 (0.96 to 1.00)
Negative	2	1,703

^aComposite reference method (CRM) was based on the results from two PCR/bidirectional sequencing assays. A true positive was positive by at least one bidirectional sequencing assay. A true negative was determined if both bidirectional sequencing assays were negative.

^b±, upper/lower 95%.

^cCI, confidence interval; PPA, positive percent agreement; NPA, negative percent agreement.

^dAlmost perfect (>0.90), strong (0.80 to 0.90), moderate (0.60 to 0.79), weak (0.40 to 0.59), minimal (0.21 to 0.39), or none (0 to 0.20).

^eTwo prospective urine samples were negative by CMV sequencing assay 1 and positive by CMV sequencing assay 2. The routine result from the study site lab-developed PCR was negative for CMV in both samples.

^fTwo retrospective urine samples were negative by the CRM. The study site’s lab-developed PCR result was positive for both samples. The Simplexa Congenital CMV Direct assay had a cycle threshold (C_T) of 34.8 and a C_T of 36.8 for the two samples, respectively.

LoD of saliva swab specimens.

Using quantified stocks of three CMV strains (AD-169, Towne, and Merlin) serially diluted in CMV-negative saliva, the LoD by positivity rate greater than 95% in 48 replicates was determined to be 5 copies/reaction in the three transport systems assessed, namely, UTM with 1 mL of transport medium, UTM with 3 mL of transport medium, and M4RT with 3 mL of transport medium (Table 3). The LoD by 95% probit analysis in the three transport systems was between 3 to 5 copies/reaction, corresponding to 274 to 465 copies/mL or 352 to 480 international units (IU)/mL.

TABLE 3.

Summary of limit of detection results for Simplexa Congenital CMV Direct-saliva swab

	Positive rate (%)a					Probit (±95% CI)a,b
Strains	cp/mL in UTM (1 mL)	IU/mL in UTM (1 mL)	cp or IU/Rx	Percent detection (no. detected/no. tested)	Mean CT ± SD (% CV)	cp/mL in UTM (1 mL)	IU/mL in UTM (1 mL)	cp/Rx
AD-169	500	NA	5	100.0% (48/48)	36.6 ± 0.9 (2.5%)	274 ± 12	NA	3
Towne	500	NA	5	95.8% (46/48)	36.5 ± 0.9 (2.6%)	465 ± 12	NA	5
Merlin	NA	500	5	95.8% (46/48)	37.7 ± 1.2 (3.3%)	NA	480 ± 12	5
Strains	cp/mL in UTM (3 mL)	IU/mL in UTM (3 mL)	cp or IU/Rx	Percent detection (no. detected/no. tested)	Mean CT ± SD (% CV)	cp/mL in UTM (3 mL)	IU/mL in UTM (3 mL)	cp/Rx
AD-169	500	NA	5	100.0% (48/48)	36.0 ± 1.0 (2.8%)	300 ± 14	NA	3
Towne	500	NA	5	100.0% (48/48)	35.4 ± 1.6 (4.4%)	346 ± 10	NA	3
Merlin	NA	500	5	100.0% (48/48)	37.3 ± 1.0 (2.7%)	NA	363 ± 14	4
Strains	cp/mL in M4RT (3 mL)	IU/mL in M4RT (3 mL)	cp or IU/Rx	Percent detection (no. detected/no. tested)	Mean CT ± SD (% CV)	cp/mL in M4RT (3 mL)	IU/mL in M4RT (3 mL)	cp/Rx
AD-169	500	NA	5	100.0% (48/48)	36.3 ± 0.7 (2.0%)	360 ± 12	NA	4
Towne	500	NA	5	100.0% (48/48)	35.4 ± 1.1 (3.0%)	315 ± 12	NA	3
Merlin	NA	500	5	95.8% (46/48)	37.3 ± 1.0 (2.7%)	NA	352 ± 12	4

^aCI, confidence interval; cp, copies; IU, international units; NA, not applicable; Rx, reaction; SD, standard deviation; CV, coefficient of variation.

^b±, upper/lower 95%.

LoD of urine specimens.

The LoD by positivity rate greater than 95% in 48 replicates of each of the stock CMV strains in CMV-negative urine was 4 copies/reaction for AD-169, 8 copies/reaction for Towne, and 64 IU/reaction for Merlin (Table 4). The LoD by 95% probit analyses was 4 copies/reaction (350 ± 12 copies/mL) for AD-169, 7 copies/reaction (652 ± 11 copies/mL) for Towne, and 40 IU/mL (3,997 ± 58 IU/mL) for Merlin.

TABLE 4.

Summary of limit of detection results for Simplexa Congenital CMV Direct-urine

	Positive rate (%)a					Probit (±95% CI)a,b
Strains	cp/mL	IU/mL	cp/Rx	Percent detection (no. detected/no. tested)	Mean CT ± SD (% CV)	cp/mL	IU/mL	cp/Rx
AD-169	400	NA	4	100.0% (48/48)	37.5 ± 1.3 (3.5%)	350 ± 12	NA	4
Towne	800	NA	8	100.0% (48/48)	35.7 ± 0.8 (2.3%)	652 ± 11	NA	7
Merlin	NA	6,400	64	100.0% (48/48)	35.4 ± 0.6 (1.8%)	NA	3,997 ± 58	40

^aCI, confidence interval; cp, copies; IU, international units; NA, not applicable.

^b±, upper/lower 95%.

Paired samples and distribution of C_T values.

A total of 122 subjects with matched saliva and urine samples were tested using the Simplexa Congenital CMV Direct assay. Of these, 120 were negative for CMV in both urine and saliva. One of the subjects (19-day-old female) with paired samples had CMV-positive saliva (C_T = 35.5) and urine (C_T = 32.5), and the other (14-day-old female) had a saliva C_T value of 23.8 and a urine C_T value of 22.6. Among all CMV-positive saliva and urine samples in this study (paired and unpaired), the means (25.8 for saliva and 26.4 for urine) and distributions of C_T values were not statistically different (Fig. 1).

FIG 1.

Distribution of Simplexa Congenital CMV Direct C_T values from CMV-positive saliva swabs and urine samples.

DISCUSSION

The incidence of cCMV infection is greater than the incidence of all other disorders combined that are included in the universal newborn screening program (17). Yet, most pregnant women have not heard of CMV infection and its associated sequelae (18). Healthcare providers are not likely to have complete information on cCMV and do not routinely provide education or counseling for pregnant women (19). Several states in the United States have now mandated CMV testing for newborns that fail a hearing screen within the first 3 weeks of life and, in some cases, require education of the public and health care professionals about cCMV (20). Most recently, with the passage of the Vivian Act in February 2022, Minnesota became the first state in the United States that will screen all newborns for congenital CMV (21). The recommended testing is typically PCR to detect CMV in the saliva or urine of newborns (13, 22).

Both urine and saliva typically have high concentrations of CMV in congenitally infected newborns, generally in the range of 10⁴ to greater than 10⁷ IU/mL (11, 23). A large prospective study of cCMV screening showed that CMV PCR in saliva has excellent sensitivity and specificity compared to rapid saliva culture (24). Some small studies reported similar performance between urine and saliva (25, 26). However, false-positive cCMV diagnoses using only saliva testing would be an issue because CMV is frequently present in breastmilk and the birth canal. Therefore, contamination of the newborn saliva samples could happen throughout the peripartum period (27). For this reason, newborn saliva should be collected more than 1 h after breastmilk consumption, and it is recommended to collect and test a urine specimen to confirm the saliva result (11, 13). The Simplexa Congenital CMV Direct real-time PCR test is FDA cleared for use with both saliva and urine specimens, which allows for screening and confirmation with a single assay on one platform with results available about 1 h from the start of testing.

Collection of a newborn saliva swab specimen can be done relatively easily and is a common source for testing newborns for cCMV (11, 24, 25). Testing with the Simplexa Congenital CMV Direct assay on saliva swabs showed a 98.6% PPA and 99.9% NPA with a κ of 0.98 (almost perfect) compared to the composite reference method (Table 1). The single false negative was found to be positive by only one of two PCR/bidirectional sequencing assays and was also negative by the study site lab-developed PCR for CMV. One Simplexa Congenital CMV Direct false positive (C_T = 37.0) was not detected by the PCR/bidirectional sequencing assays but was positive by the study site lab-developed PCR. Both of these samples may represent those from rare patients with very low viral loads (i.e., high C_T values) that can be observed in cCMV (23, 28), which are not reproducibly detected by different PCR assays perhaps due to subtle differences in detection at or near the limit of detection of the respective assays. Low-level viral loads have been found in both true-positive and false-positive cases of cCMV, and, due to the poor positive predictive value with saliva testing alone, a second sample (typically urine) is recommended for confirmation (29). Although obtaining a urine specimen from neonates suspected of cCMV may be more challenging than collection of a saliva swab, testing of urine by PCR is considered the “gold standard” for diagnosing cCMV in infants less than 21 days old, particularly compared to previously used culture methods (30). One CMV PCR-negative urine sample in a neonate is usually sufficient to exclude infection, and the Simplexa Congenital CMV Direct assay demonstrated 99.8% overall accuracy with this specimen type (Table 2).

The analytical sensitivity of the Simplexa Congenital CMV Direct assay for saliva swabs in 1 mL or 3 mL of transport medium was found to be less than 500 IU/mL (Table 3), a threshold significantly lower than virus concentrations typically found in newborns with cCMV (23) and a substantially lower limit of detection than the other FDA-cleared assay for saliva testing (31). Likewise, the analytical sensitivity for urine specimens using the Simplexa Congenital CMV Direct test was less than 2.8 log DNA copies/mL, which is considerably lower than the median urine viral load of 6.1 log DNA copies/mL found in one study of newborns with cCMV (11) (Table 4). Specifically, probit analysis demonstrated a limit of detection of 350 ± 12 copies/mL for the AD-169 strain, 652 ± 11 copies/mL for the Towne strain, and 3,997 ± 58 IU/mL for the Merlin strain. These data are comparable to those from another independent study of the analytical sensitivity of the Simplexa Congenital CMV Direct assay (28).

Although no specimens used in this study had C_T values above the FDA-cleared parameters, which define a “detected” result only for those with values ≤37.5, the analytical sensitivity studies demonstrated that CMV can be detected with C_T values of ≥37.6 (Tables 3 and 4). A recent study also showed that 2 of 34 CMV-positive urine samples would have been classified as “not detected” by the FDA-cleared criteria with the Simplexa Congenital CMV Direct assay, although amplification curves could be visualized with C_T values of 38.3 and 40.3 available in a separate tab of the software (28). In these instances, laboratories might consider follow-up testing to confirm the presence or absence of CMV in such samples.

The major strengths of this study are the inclusion of a large number of prospectively collected newborn saliva swabs (n = 1,853) and urine specimens (n = 1,624). Additionally, the overall number of CMV-positive saliva (n = 70) and urine samples (n = 92), as determined by the composite reference method, allowed for a high-level statistical assessment of assay performance. The limitations include a lack of patient clinical data to determine if newborns had symptoms compatible with cCMV and the inability to calculate positive and negative predictive values for the condition. Additional studies with a larger number of paired urine and saliva swab specimens are warranted.

In conclusion, our study showed that the Simplexa Congenital CMV Direct assay could be applied to both saliva and urine specimens collected from newborns less than 21 days of age to effectively rule in or rule out congenital CMV infection in a rapid, sensitive, and specific manner. In most instances, it is recommended that a CMV-positive saliva test be confirmed by urine testing (13, 32), so the availability of this FDA-cleared test allows institutions to easily use screening and confirmatory testing on site for the most common congenitally acquired infection.