Simultaneous Detection and Differentiation between Wild-Type and Vaccine Measles Viruses by a Multiplex Real-Time Reverse Transcription-PCR Assay

Measles is one of the most contagious viral respiratory infections and was declared to be eliminated from Canada in 1998; however, measles cases and outbreaks still occur every year through reintroduction from other parts of the world. Laboratory confirmation of measles virus (MV) RNA by real-time PCR provides a definitive diagnosis, and molecular analysis to determine the genotype is the only way to distinguish between wild-type and vaccine strains.

ABSTRACT

Measles is one of the most contagious viral respiratory infections and was declared to be eliminated from Canada in 1998; however, measles cases and outbreaks still occur every year through reintroduction from other parts of the world. Laboratory confirmation of measles virus (MV) RNA by real-time PCR provides a definitive diagnosis, and molecular analysis to determine the genotype is the only way to distinguish between wild-type and vaccine strains. This distinction is important since live attenuated vaccine strains are able to replicate in the patient and can be associated with rash and fever but are poorly transmissible, if at all. Prompt reporting of measles cases to local authorities, including differentiation between wild-type and vaccine strains, allows for optimal management and contact tracing. The development and validation of a multiplex real-time reverse transcription-PCR (rtRT-PCR) assay for the simultaneous detection and differentiation of the Moraten and Schwarz vaccine strains from presumptive wild-type MV in a format that can be easily implemented for high-throughput testing of patient samples are reported here. This assay is sensitive, specific, reproducible, and 100% accurate in comparison with the gold standard comparator assay.

INTRODUCTION

Measles is an acute febrile illness characterized by a high fever, conjunctivitis, coryza, cough, and a maculopapular rash. It is highly contagious and is spread by contact with droplets and aerosols containing the virus (1). The infectious period lasts from several days before to several days after the onset of rash; this also coincides with the most intense period of cough and coryza, facilitating transmission (2). The highly contagious nature of the measles virus (MV) is reflected by its basic reproductive number (R₀), (3, 4). For MV the R₀ value ranges from 15 to 17, whereas the R₀ values of mumps virus and influenza virus (2, 5), which are also agents associated with outbreaks or epidemics, are 10 to 12 and 2 to 3, respectively. As a result, vaccination coverage rates of at least 95% are required to interrupt transmission and eventually achieve measles elimination (6, 7). The interruption of endemic measles was achieved in Canada by 1998 (8), 2 years later in the United States (9), and by November 2002 in the Americas (6) through comprehensive immunization programs. However, the elimination of measles is not a static state and requires significant ongoing vaccination efforts to maintain it. In contrast, measles continues to circulate in Europe, Asia, and Africa, as well as in the Americas, in 2018, as evidenced by outbreaks and sporadic cases reported to the WHO (http://www.who.int/immunization/monitoring_surveillance/burden/vpd/surveillance_type/active/measles_monthlydata/en/). This appears to be mainly the result of antivaccination attitudes (10), and the trend is especially worrisome because even in countries with well-vaccinated populations there remain susceptible subgroups. Vulnerable populations include children less than 1 year of age, the immunocompromised, and those with uncertain vaccination status, who may travel to or be exposed to infectious visitors from areas of endemicity. Canada is a popular tourist destination and, together with its multiethnic population, attracts both tourists and visiting family members from countries where measles is endemic. Thus, imported sporadic measles cases resulting in secondary spread of the virus to susceptible individuals are always a possibility. This is particularly concerning when immunodeficient patients, such as transplant recipients, are exposed to infectious measles cases, for example, in a health care setting, as it can cause severe and often fatal illnesses, such as giant cell pneumonitis or measles inclusion body encephalitis (MIBE), in these individuals (1).

The administration of a measles-containing vaccine can result in a febrile rash-like illness with a presentation similar to that of mild measles in up to 5% of recipients (11), and it is difficult to distinguish vaccine-attributable cases from sporadic wild-type infections. A clinical diagnosis can also be difficult in individuals who may not have typical symptoms due to the presence of preexisting antibodies from maternal immunity or immune globulin or from a previous vaccination, a longer incubation period, or the presence of a milder prodromal illness or rash (12). Requests for measles testing can be triggered on the basis of exposure during travel, if a recent vaccination history is not easily accessible, for active public health surveillance initiatives, or if other infectious agents, such as enterovirus, the features of infection with which can mimic some features of a measles infection, are circulating. In Alberta and other Canadian provinces, active public health surveillance for measles and rubella requires health care practitioners to report suspect cases for further investigation, which often includes laboratory testing for patients with a febrile illness and rash.

The cost and human resources required for contact tracing and management of suspected measles cases can be substantial; hence, the availability of local, rapid, and definitive testing can be a significant cost-saving strategy. Such testing also facilitates the prompt isolation of infectious cases using airborne precautions to interrupt transmission. Vaccine strains are poorly or not transmissible, and prompt differentiation between wild-type and vaccine strains is required to allow optimal management and public health action (12–15).

The development and validation of a triplex real-time reverse transcription-PCR (rtRT-PCR) assay for the simultaneous detection and discrimination between vaccine and presumptive wild-type MV strains are reported here. The triplex assay includes three independent probes with unique reporter dyes, used for (i) detection of MV, (ii) detection of the Schwarz and Moraten vaccine strains, and (iii) detection of wild-type MV. Primers and a probe for the detection of all genotypes of MV, including vaccine strains, target the conserved RNA polymerase region (L gene); an additional set of primers and single nucleotide polymorphism (SNP)-specific probes detecting the hemagglutinin (H) gene targets a single nucleotide polymorphism unique to the Schwarz and Moraten strains, which are the vaccine strains used in North America. In Canada, four vaccines are licensed for use, M-M-R II and ProQuad, which are manufactured by Merck and which use the Moraten strain, and Priorix and Priorix-Tetra, which are manufactured by GlaxoSmithKline (GSK) and which use the Schwarz strain; thus, the assay addresses both the postvaccination and wild-type scenarios that can be encountered in the population (16, 17; https://www.canada.ca/en/public-health/services/publications/healthy-living/canadian-immunization-guide-part-4-active-vaccines/page-12-measles-vaccine.html).

MATERIALS AND METHODS

Design of primers and probes.

Representative sequences (as of 7 December 2016) of the L gene from different MV genotypes in the GenBank nonredundant database were aligned for the design of primers and a hydrolysis probe that had a minor groove binding (MGB) protein and that was labeled with 6-carboxyfluorescein (FAM) as the fluorescent reporter dye. The hydrolysis probe was purchased from Applied Biosystems (ABI; Foster City, CA). H gene sequences from genotypes A, B1, B2, B3, C1, C2, D1, D2, D3, D4, D5, D6, D7, D8, D9, D10, D11, E, F, G1, G2, G3, H1, and H2, vaccine strains derived from the Edmonston strain, including AIK-C, Rubeovax, Zagreb, Moraten, and Schwarz, and the wild-type Edmonston strain were aligned for the design of primers and probes for the differentiation of wild-type and vaccine strains. Two locked nucleic acid (LNA) probes (Table 1) were designed for the differentiation of the wild-type and the Schwarz and Moraten vaccine strains, taking advantage of an SNP unique to the Schwarz and Moraten vaccine strains compared to the sequences of other vaccine strains and wild-type viruses (16, 17). The probe for the detection of the Schwarz and Moraten vaccine strains was labeled with Cy5 as the fluorescent reporter dye and Iowa Black as the nonfluorescent quencher; additionally, a TAO internal quencher was used to avoid cross detection of the Cy5 signal in other fluorescent channels. The probe for the detection of wild-type strains was labeled with HEX and Iowa Black. The LNA probes were purchased from Integrated DNA Technologies (IDT; San Diego, CA). Primers flanking the detection region were designed for the amplification of a longer fragment to generate plasmid clones used in the preparation of in vitro-transcribed RNA for the L and H genes from genotypes B3 and D8 and the Schwarz vaccine strain (obtained from an aliquot of the Priorix-Tetra vaccine [GSK]). The sequences of all the oligonucleotides used are provided in Table 1. The primers and probes were designed using Primer Express (v3.0) software (ABI), and all sequence alignments were performed using the ClustalW program (BioEdit). All primers were purchased from the University Core DNA Services (University of Calgary, Calgary, Alberta, Canada).

TABLE 1.

Primers and probes for the multiplex assay

Primer purpose and type	Name	Sequence (5′-3′)a
Measles virus detection
Forward	Measles_L_For	GCATCGAGAGAGGTTATGACCG
Probe	Measles_L_FAM_MGB	FAM-CTTGGCTTCACAATCA-NFQ
Reverse	Measles_L_Rev	TTGTGAGGAGGGGTATGACTACATC
Cloning forward	Measles_L_CloneF	GCTACTTGTGTCCCAATCACTC
Cloning reverse	Measles_L_CloneR	TCACCGATGTTTCTGACAAACA
Measles virus typing
Forward	Measles_H_314For	TGAGGACACCTCAGAGATTCACTG
Reverse	Measles_H_414Rev	CCAAGTGAGATCTCTGAAGTCGTACTC
Probe_wtb	Measles_H_LNAwt_HEX	HEX-TG+AAAT+T+C+AT+CT+CTG-IABkFQ
Probe_vaccine	Measles_H_LNAvac_Cy5	Cy5-TGAAAT+T+A+A-TAO-T+CT+CT+GAC-IABkFQ
Cloning forward	Measles_H_52For	GGGAAGTAGGATAGTTATTAACAGAGAAC
Cloning reverse	Measles_H_662Rev	GTGACTATAGATGACACATTGTAACC

^a+, an LNA base. The fluorescent reporter dyes (FAM, HEX, and Cy5) and internal (TAO) and terminal (NFQ, nonfluorescent quencher; IABkFQ, Iowa Black dark quencher) quenchers are indicated. The single nucleotide polymorphism differentiating between wild-type and vaccine strains is underlined and bold.

^bwt, wild type.

The gold standard assay used for comparison of the newly developed triplex assay was adapted from a method previously described by the Centers for Disease Control and Prevention (CDC) (18). This is a hydrolysis probe-based rtRT-PCR assay targeting the hemagglutinin (H) and nucleoprotein (N) genes performed on a LightCycler real-time PCR system platform (the LC assay; Roche). With the 2013 and 2014 Canadian outbreaks of measles virus genotype B3 strain Harare (identical to MVi/Harare.ZWE/38.09/1 [GenBank accession number JF973033]), this assay gave a negative reverse transcription-PCR (RT-PCR) assay result from the H gene target because of mismatches in the probe region, despite positive results with the N gene target. The National Microbiology Laboratory (NML; Winnipeg, Manitoba, Canada) modified this assay by introducing an additional probe capable of hybridizing to the circulating B3 genotype strains (probe MVH154p-B3 [FAM-TTG CTG GCA ATT GCA GGC ATT-Blackberry nonfluorescent quencher], designed at NML). (19). This protocol was implemented at the Provincial Laboratory for Public Health, Calgary, Alberta, Canada, in 2016.

Real-time RT-PCR assay.

A one-step RT-PCR was performed using a TaqMan Fast virus one-step RT-PCR master mix (ABI), 0.8 μM each sense and antisense primers, and 0.2 μM probes. Five microliters of the extracted RNA was combined with 5 μl of the master mix, and the reverse transcription step was performed at 50°C for 5 min, followed by incubation at 95°C for 20 s. Amplification included 45 cycles of denaturation at 95°C for 3 s, followed by annealing, extension, and data acquisition at 60°C for 30 s on a 7500 Fast real-time PCR system (ABI).

Preparation of RNA transcripts for sensitivity studies.

The cloning primers listed in Table 1 were used for the amplification of longer PCR products from the vaccine strain and wild-type MV genotypes B3 and D8. These products were cloned using a TOPO TA cloning dual promoter kit (Life Technologies, CA, USA). RNA transcription was performed using a RiboMAX SP6 or T7 RiboMAX Express RNA production system (Promega, Madison, WI, USA) and standard protocols. The transcribed RNA was spectrophotometrically quantified for the calculation of copy numbers.

Extraction of viral nucleic acid.

Viral RNA was extracted from nasopharyngeal swab and urine samples using an easyMAG automated extractor (bioMérieux, Quebec, Canada) according to the manufacturer’s instructions. The sample input volume was 200 μl, and the output volume was 110 μl.

Analytical sensitivity, dynamic range, specificity, and reproducibility of RT-PCR.

Tenfold serial dilutions of RNA from genotypes B3 and D8 and the vaccine strain quantified in vitro were used to determine the analytical sensitivity by testing in triplicate in three independent runs. The 95% limits of detection (LOD 95%) were calculated by probit analysis using Microsoft Excel software, followed by rounding up of the copy number. The viral loads tested to determine the linear dynamic range of the assay for the H and L genes from the wild-type and vaccine strains are indicated in Table 2. Linear regression fitting of the log viral load versus the threshold cycle (C_T) value allowed the calculation of C_T values corresponding to the LOD 95%.

TABLE 2.

Assay characteristics

Strain	L gene detection				H gene typing
	Dynamic range (no. of copies/reaction)	LOD 95% (no. of copies/reaction)	CT corresponding to LOD 95%a	Efficiency (%)	Dynamic range (no. of copies/reaction)	LOD 95% (no. of copies/reaction)	CT corresponding to LOD 95%	Efficiency (%)
GSK vaccine	1.24E1–1.24E6	2	36.90	100.51	1.01E1–1.01E7	7	38.16	117.25
B3	6.02E1–6.02E6	10	37.19	111.17	4.14E1–4.14E7	9	39.57	101.66
D8	1.40E1–1.40E6	2	36.87	101.17	8.48E2–8.48E8	1,206	38.00	102.81

^aLinear regression plots of the copy number and C_T values were used to calculate the C_T value corresponding to the LOD 95%.

The pathogens used for the testing of specificity included those that could cause symptoms mimicking those of a measles virus infection and those commonly encountered in a nasopharyngeal swab specimen, which is the specimen of choice for the direct detection of MV. Samples with high viral or bacterial loads for the following pathogens were used: rubella virus, human herpesviruses 6 and 7, influenza A and B viruses, respiratory syncytial virus A, parainfluenza virus 3, human rhinovirus 1b, echovirus 2, coxsackievirus A16, human metapneumovirus, adenovirus serotype 4, bocavirus, human coronavirus HKU1, parvovirus B19, Bordetella pertussis, Haemophilus influenzae, and Streptococcus pneumoniae. The reproducibility of the triplex assay was evaluated using specimens positive for vaccine strains of MV belonging to genotypes A (C_T = 32.41), B3 (C_T = 25.73), and D8 (C_T = 27.28). All three specimens were tested in triplicate in four independent runs, and the C_T values were used to calculate the interassay and intra-assay variability.

Clinical specimens and testing of different genotypes.

Available specimens that had been submitted to the Provincial Laboratory from 2013 to 2017 for the investigation of MV and that had been previously tested by the gold standard assay from CDC (18) were retested by the triplex assay.

A total of 104 samples (positive, n = 76; negative, n = 28), including respiratory specimens (n = 62), urine specimens (n = 40), a cerebrospinal fluid specimen (n = 1), and a brain tissue specimen from a subacute sclerosing panencephalitis case (n = 1) (20), were tested for MV and differentiated as a wild-type strain or a vaccine strain. The positive samples were genotyped at NML by sequencing 450 bp of the partial nucleoprotein gene in the carboxy-terminal region and included samples with genotype A (n = 21), B3 (n = 29), D6 (n = 1), D8 (n = 17), and H1 (n = 4) viruses and untypeable viruses (n = 4). All samples with untypeable viruses had a very low viral load and, thus, failed to generate a sequence. The results of the gold standard assay and the genotyping results were compared to the results of the newly developed triplex assay.

Tenfold serial dilutions of MV cultures of genotypes B2, C1, C2, D2, D3, D4, D5, D7, D9, D10, G1, G2, and H2 were obtained from NML and tested by the triplex assay and the gold standard LightCycler assay. In addition, 10-fold dilutions of the genotype A-derived vaccine strains M-M-R II (Moraten strain; Merck) and Priorix (Schwarz strain; GSK) were tested by the gold standard method and the triplex assay. Serial dilutions of patient samples available for genotypes D6, H1, B3, and D8 were tested simultaneously by both methods. Genotypes B1, D1, D11, F, and G3 were not available for testing and were evaluated in silico for the H gene; the L gene sequence is currently not available for genotypes B1, D1, D11, and F.

The results of the triplex assay were scored as follows: “negative” if both the L gene and the H gene SNP probes were negative, “Schwarz or Moraten vaccine strain identified” if the L gene PCR was positive and the SNP probe was positive for the vaccine strain, and “presumptive wild-type measles virus identified” otherwise. Other vaccine strains, including Edmonston-derived vaccine strains, would not be identified by the vaccine probe; these other vaccine strains are, however, not licensed for use in Canada.

RESULTS

Assessment of RT-PCR assay performance: analytical sensitivity, specificity, and reproducibility.

The results for analytical sensitivity, dynamic range, and assay efficiency are summarized in Table 2. The LOD 95% by the triplex assay was 2, 10, and 2 copies/reaction for the L gene assay for the vaccine strains and strains of the B3 and D8 genotypes, respectively. The LOD 95% using the probe targeting the H gene from the vaccine strain was 7 copies/reaction, and that for the probe targeting the H gene from the wild-type strain was 9 and 1,206 copies/reaction for the B3 and D8 genotypes, respectively. The C_T value corresponding to the LOD 95% was calculated for each of the genotypes with the three probes. These values ranged from 36.87 to 39.57 (Table 2). Based on these values, a C_T cutoff of 35 was selected as the threshold; samples with C_T values of less than 35 were reported as positive, and samples with C_T values of greater than 35 were retested in duplicate for confirmation.

The triplex assay did not amplify any of the other pathogens tested, thus showing specificity for MV.

The intra-assay variability (percent coefficient of variation [CV]) was calculated using the replicates within the same run and varied from 0.16% to 1.22%. The interassay variability was calculated using values obtained from the different runs and ranged from 0.20% to 5.13%, showing reproducible detection (Table 3).

TABLE 3.

Intra- and interassay variability at different viral loads for the vaccine and commonly circulating wild-type strains

Gene or strain (dye or quencher), genotype	Original CT	Intra-assay run 1			Intra-assay run 2			Intra-assay run 3			Intra-assay run 4			Interassay variability
		Avg CT	CT SD	% CV	Avg CT	CT SD	% CV	Avg CT	CT SD	% CV	Avg CT	CT SD	% CV	Avg CT	CT SD	% CV
L gene (FAM)
A	32.41	33.52	0.16	0.49	31.78	0.32	1.01	32.65	0.29	0.88	32.33	0.23	0.70	32.57	0.73	2.24
B3	25.73	25.50	0.04	0.16	24.29	0.18	0.76	25.19	0.06	0.24	25.25	0.09	0.36	25.06	0.53	2.11
D8	27.28	27.08	0.29	1.06	25.77	0.12	0.46	26.78	0.10	0.36	26.64	0.06	0.21	26.57	0.56	2.12
MV vaccine strain (Cy5-TAO)	34.83	35.94	0.14	0.38	35.95	0.41	1.14	32.19	0.14	0.42	34.47	0.42	1.22	34.64	1.78	5.13
MV wild type (HEX)
B3	31.09	30.01	0.09	0.28	29.97	0.17	0.55	29.77	0.16	0.53	29.71	0.21	0.71	29.87	0.15	0.49
D8	35.45	32.57	0.15	0.45	32.46	0.22	0.69	32.55	0.10	0.29	32.44	0.25	0.76	32.51	0.06	0.20

Testing of clinical samples.

At the Provincial Laboratory, 104 samples submitted for MV PCR were available for comparison. Of these, 28 samples negative by the LC method were also negative by the triplex assay. Sixty-two samples positive by the LC assay (H and N gene PCRs) were tested by the triplex assay; of these, 19 samples that typed as genotype A at NML (i.e., vaccine strains) were positive by the triplex assay and correctly identified as Schwarz or Moraten vaccine strains. Forty-one samples positive by the LC assay were positive and identified as presumptive wild-type viruses by the triplex assay; these included 18 samples belonging to genotype B3, 17 to genotype D8, 1 to genotype D6, and 4 to genotype H1, and one sample was untypeable. Additionally, two samples positive by the LC assay had a very low viral load and so the viruses were untypeable. The first one had a C_T value of 35.72 by the assay for the H gene and a C_T value of 37.82 by the assay for the N gene and tested positive as a vaccine strain by the triplex assay in 1 of 3 replicates, with a C_T value of 35.24 by the assay for the L gene and a C_T value of 37.33 by the assay with the vaccine probe. The second sample had C_T values of 39.28, 36.13, and negative in 3 replicates of the assay for the H gene and C_T values of negative, 37.66, and >40 in 3 replicates of the assay for the N gene and was positive by the L gene PCR in the triplex assay in 2 of 3 replicates (C_T values, 35.05 and 34.91).

The results for 14 samples were initially reported as “inconclusive” by the LC assay (i.e., positive by the H gene target PCR but negative by the N gene target PCR). Of these, 12 had genotype B3 and a recently described mutation that prompted a modification of the N gene PCR (as discussed above). Viruses were detected in 11 of these samples and typed as presumptive wild type by the triplex assay. The three remaining samples, of which two were untypeable, had a low viral load and were positive only by the L gene PCR using the triplex assay.

Thus, for all isolates that were positive for both targets by the gold standard assay (LC assay) and for which a genotype was available, the triplex assay was also positive and correctly differentiated vaccine strains from presumptive wild-type isolates. There was also complete agreement for negative samples.

Sensitivity for different genotypes.

Tenfold serial dilutions of cultures and patient samples for all available genotypes listed in the Materials and Methods section were tested simultaneously by the gold standard method and the triplex assay. The results are shown in Table 4. The sensitivities of the gold standard and triplex assay were within 1 order of magnitude for all the genotypes tested, and the triplex assay correctly identified the vaccine and wild-type strains.

TABLE 4.

Comparison of sensitivity for serial dilutions of different genotypes^a

Genotype (dilution)	Triplex TaqMan CTb			LC (Cp)
	L gene	MV vaccine strain	MV wild-type strain	H gene	N gene
Quantitative results
B2 (10−2)	32.16	Neg	35.40	36.22	36.17
B2 (10−3)	35.02	Neg	Neg	Neg	Neg
C1 (10−3)	30.68/31.20	Neg	32.45/32.64	32.41	32.52
C1 (10−4)	Neg/Neg/Neg	Neg/Neg/Neg	Neg/Neg/Neg	36.50	Neg
C2 (10−3)	33.42	Neg	36.88	35.28	34.56
D2 (10−3)	32.72	Neg	35.30	35.07	36.06
D3 (10−5)	33.32	Neg	36.74	36.13	35.86
D3 (10−6)	Neg	Neg	Neg	36.09	Neg
D4 (10−3)	32.8	Neg	36.94	32.66	29.99
D4 (10−4)	Neg	Neg	Neg	36.31	36.78
D5 (10−3)	33.94	Neg	35.22	34.90	37.93
D6 (10−4)	31.72	Neg	31.45	35.67	29.42
D6 (10−5)	33.84	Neg	33.97	Neg	32.38
D7 (10−2)	33.47	Neg	35.29	35.60	37.52
D9 (10−4)	34.38	Neg	36.87	35.60	Neg
D10 (10−2)	42.74/40.08	Neg	35.03/35.94	34.92	35.99
G1 (10−2)	29.81	Neg	31.48	32.17	32.14
G1 (10−3)	34.58	Neg	Neg	Neg	38.63
G2 (10−2)	35.04	Neg	34.89	36.06	35.93
H1 (10−1)	33.21	Neg	35.99	35.63	37.07
H1 (10−2)	Neg	Neg	Neg	Neg	>40
H2 (10−2)	31.8	Neg	35.11	35.85	36.80
H2 (10−3)	Neg	Neg	39.27	Neg	Neg
A_GSK_vac (10−3)	34.07	35.15	Neg	34.95	35.26
A_GSK_vac (10−4)	35.12	Neg	Neg	>40	>40
A_Merck_vac (10−2)	31.54	34.36	Neg	33.60	34.14

^aTenfold serial dilutions of cultures or genotyped patient samples belonging to different genotypes were compared by both assays. C_p, crossing point on the Light Cycler platform; Neg, negative.

^bC_T values from samples tested in replicates are separated by a slash.

DISCUSSION

Even though MV cases continue to be diagnosed, the number of wild-type MV infections has considerably diminished with vaccination, and the relative frequency of cases where a vaccine strain is detected has increased. In Alberta during the period from 2010 to 2018, 30% of cases where a genotype could be determined were caused by a vaccine strain.

It is important to distinguish wild-type-related disease from vaccine-related disease since the failure to quickly identify a vaccine strain can lead to unnecessary quarantine, contact tracing, and measles prophylaxis in a newborn (12, 15, 21). The assay reported here can be used for the simultaneous detection and typing of MV, facilitating timely public health decisions to stop the transmission cycle and undertake appropriate management and treatment strategies. This is especially critical because of the high infectivity of wild-type MV (2).

Molecular markers that allow for the differentiation between wild-type and vaccine strains have been previously identified in different genes (17, 22). An MV genotype A-specific reverse transcription-quantitative PCR assay targeting the N gene has been published (23–25), as have a number of real-time RT-PCR assays for the detection of MV (26–29). In this study, these strategies were multiplexed in order to simultaneously detect and differentiate wild-type MV from the vaccine strains using two independent gene targets. Since the assay can be performed using fast cycling conditions on an ABI 7500 instrument, it is amenable to high-throughput testing in a diagnostic laboratory with a short turnaround time. The assay is validated for respiratory specimens and urine, which are the recommended sample types for MV testing.

WHO has reported that of the 24 known MV genotypes, only B2, B3, D4, D5, D6, D7, D8, D9, D11, G3, and H1 were circulating globally between 2005 and 2014, and of these, H1, B3, D8, D9, D4, and G3 were reported in that order of prevalence; however, this may not accurately indicate their global distribution because of reporting bias (30). Genotypes B3 and D8 have been reported to be the most widely distributed types (30). Genotypes C2, D2, D3, G2, and H2 are now considered inactive by WHO, as there were no reports of these genotypes circulating for at least 10 years prior to 2015 (30). The detection of all circulating genotypes was confirmed by the triplex assay using either virus culture or patient samples for all genotypes except genotypes D11 and G3, which were not available at the Provincial Laboratory. Of note, only 7 cases of genotype D11 infection and 103 cases of G3 infection were reported worldwide between 2005 and 2014 (30), based on information in the MeaNS (Measles Nucleotide Sequence) database (http://www.who-measles.org/), which collects sequence information submitted mainly by members of the Global Measles and Rubella Lab Network (GMRLN) or downloaded from GenBank. In silico comparison of the sequence of the L gene from genotype G3 showed a single base mismatch with the forward primer sequence close to the 5′ end and a perfect match with the sequences of the probe and reverse primer; no L gene sequences for genotype D11 were available for comparison. In silico comparison of the H gene from genotype D11 showed a perfect match with the probe and forward primer sequences, and the reverse primer had one base mismatch. The sequence of the H gene from genotype G3 was a perfect match with the forward primer and probe sequences, although the reverse primer had one base mismatch closer to the 5′ end, and this mismatch was also present in genotype G1, G2, and D9 strains, all of which were tested during validation of this assay and found to be reliably detected. Based on the findings of in silico analysis, it is predicted that isolates belonging to these genotypes would be reliably detected.

Thorough measurements of the sensitivity were performed for genotypes A, B3, and D8 using quantitated in vitro-transcribed RNA for the L and H genes. The sensitivity for all targets was less than 10 copies/reaction for all genotypes except genotype D8 using the H gene probe. However, since the sensitivity for genotype D8 using the L gene probe was 2 copies/reaction and the H gene assay result is used for differentiation between vaccine and wild-type strains, the assay would be positive for MV but may not provide strain differentiation in samples with a very low viral load; in such cases, it may be prudent to assume that a wild-type virus is present, which is how the result would be reported. However, as the viral load in an infected individual in the prodromal phase or in an individual with a rash presentation for up to 10 days is orders of magnitude greater than the assay’s threshold, this difference appears to be of little consequence in practice.

The operational value of this assay is the high sensitivity for all the listed circulating global genotypes coupled with the immediate differentiation between vaccine and wild-type strains with a high degree of certainty. One limitation of the SNP assay is that it is specific for the Schwarz and Moraten vaccine strains and would not identify other vaccine strains, such as Zagreb or Shanghai-191, and differentiate them from wild-type strains. However, in Canada only the Schwarz and Moraten strains are licensed for use in measles-containing vaccines, whereas in the United States, only the Moraten strain is used. Thus, in the context of samples tested at the Provincial Laboratory, the rapid identification or exclusion of the Schwarz or Moraten strain along with the patient’s history allows for optimal management.

Although the Provincial Laboratory has yet to encounter a case vaccinated outside of Canada, sequence-based genotyping for measles strain surveillance, performed at NML, would classify it as genotype A likely of vaccine origin, as wild-type genotype A strains no longer circulate.