【病毒外文文献】2004 Evaluation of Advanced Reverse Transcription-PCR Assays and an Alternative PCR Target Region for Detection of Sever

JOURNAL OF CLINICAL MICROBIOLOGY May 2004 p 2043 2047 Vol 42 No 5 0095 1137 04 08 00H110010 DOI 10 1128 JCM 42 5 2043 2047 2004 Copyright 2004 American Society for Microbiology All Rights Reserved Evaluation of Advanced Reverse Transcription PCR Assays and an Alternative PCR Target Region for Detection of Severe Acute Respiratory Syndrome Associated Coronavirus Christian Drosten 1 Lily Lily Chiu 2 Marcus Panning 1 Hoe Nam Leong 3 Wolfgang Preiser 4 John S Tam 5 Stephan Gu nther 1 Stefanie Kramme 1 Petra Emmerich 1 Wooi Loon Ng 2 6 Herbert Schmitz 1 and Evelyn S C Koay 2 6 Department of Virology Bernhard Nocht Institute for Tropical Medicine Hamburg 1 and Institute of Medical Virology Johann Wolfgang Goethe University Frankfurt 4 Germany Molecular Diagnosis Centre Department of Laboratory Medicine National University Hospital 2 Department of Infectious Diseases Tan Tock Seng Hospital 3 and Department of Pathology National University of Singapore 6 Singapore and Department of Microbiology Faculty of Medicine Prince of Wales Hospital The Chinese University of Hong Kong Hong Kong Special Administrative Region People s Republic of China 5 Received 10 October 2003 Returned for modification 14 December 2003 Accepted 14 February 2004 First generation reverse transcription PCR RT PCR assays for severe acute respiratory syndrome asso ciated coronavirus SARS CoV gave false negative results in a considerable fraction of patients In the present study we evaluated two second generation replicase R gene based real time RT PCR test kits the RealArt HPA coronavirus LC kit Artus Hamburg Germany and the LightCycler SARS CoV quantification kit Roche Penzberg Germany and a real time RT PCR assay for the nucleocapsid N gene Detecting the N gene RNA might be advantageous due to its high abundance in cells The kits achieved sensitivities of 70 8 Artus and 67 1 Roche in 66 specimens from patients with confirmed SARS samples primarily from the upper and lower respiratory tract and stool The sensitivity of the N gene assay was 74 2 The differences in all of the sensitivities were not statistically significant P H11549 0 680 analysis of variance Culture cells initially contained five times more N than R gene RNA but the respective levels converged during 4 days of virus replication In clinical samples the median concentrations of R and N gene RNA respectively were 1 2 H11547 10 6 and 2 8 H11547 10 6 copies ml sputum and endotracheal aspirates 4 3 H11547 10 4 and 5 5 H11547 10 4 copies ml stool and 5 5 H11547 10 2 and 5 2 H11547 10 2 copies sample throat swabs and saliva Differences between the samples types were significant but not between the types of target RNA All nH1154912 samples from the lower respiratory tract tested positive in all tests In conclusion the novel assays are more sensitive than the first generation tests but they still do not allow a comprehensive ruling out of SARS Methods for the routine sampling of sputum without infection risk are needed to improve SARS RT PCR Severe acute respiratory syndrome SARS involves an ini tial febrile phase followed by interstitial pneumonia which leads to respiratory distress syndrome and death in a fraction of patients The case fatality proportion has been calculated to range ca 13 2 in patients younger than 60 years of age and 43 3 in those more than 60 years old 1 The first outbreak of SARS started in southern China in autumn 2002 and spread to several countries of the northern hemisphere until it came to rest in July 2003 A novel coronavirus CoV SARS CoV has recently been identified as the causative agent of SARS 3 5 8 Irrespective of its low zero prevalence in a nonepidemic situation a diagnosis of SARS has to be considered in patients with compatible symptoms because the disease is highly con tagious and can involve large numbers of secondary cases Ideally in this situation it would be possible to rule out the disease by a laboratory test Since antibodies are detectable only late in the disease 7 virus detection by reverse transcription PCR RT PCR is most promising for this purpose Because of the low sensitivity of current methods however it is not possible to categorically rule out SARS on the basis of RT PCR results 7 10 This situation will become a major problem in case management in upcoming influenza epidemics So far only first generation in house RT PCR methods have been studied but now there are two second generation real time RT PCR test kits available both targeting the repli case R gene of SARS CoV that might have superior sensi tivity Switching the RT PCR target to the nucleocapsid N gene might increase the sensitivity even further due to the higher abundance of subgenomic N RNA in cultured cells 9 Experimental animal data support this rationale 6 but the situation in patients has not been studied yet We evaluate here both commercial tests the RealArt HPA coronavirus LC kit Artus Hamburg Germany and the LightCycler SARS CoV quantification kit Roche Penzberg Germany and an in house real time RT PCR assay for the N gene MATERIALS AND METHODS Clinical samples Sixty six samples were available from 29 retrospectively confirmed SARS patients from Singapore Hong Kong and Germany more details are given in Table 1 Twenty five patients showed seroconversion in immunofluorescence assay four patients had positive results in three indepen dent RT PCR assays equivalent World Health Organization criteria for labo Corresponding author Mailing address Bernhard Nocht Institute for Tropical Medicine Bernhard Nocht Str 74 20359 Hamburg Ger many Phone 49 40 42818 421 Fax 49 40 42818 378 E mail drosten bni hamburg de 2043 on June 17 2015 by UPVA http jcm asm org Downloaded from ratory confirmation of SARS The samples were taken after a median duration of 12 days range 2 to 54 days from the onset of symptoms The median durations from symptoms to sampling in five different categories of specimens are given in Table 1 samples from the upper and lower respiratory tract stool plasma and others These values were not significantly different from each other Kruskal Wallis test P H11005 0 092 RNA extraction RNA was extracted from the samples as described earlier 3 by using a DNA stool kit Qiagen Hilden Germany for stool specimens and a viral RNA minikit Qiagen for all other samples Sputum samples were pre treated with 2H11003 sputum lysis buffer 10 g of N acetylcysteine liter 0 9 sodium chloride for 30 min in a shaking incubator Second generation R gene RT PCR assays The RealArt HPA coronavirus LC kit and LightCycler SARS CoV quantification kit were used for amplification and quantification of the R gene according to each manufacturer s instructions Both assays yielded quantitative data in terms of RNA copy numbers per reac tion that were converted into RNA copy numbers per milliliter of liquid sample sputum stool urine plasma and saliva or total of solid sample swabs based on correction factors derived from the manipulations in the sample preparation procedures Both assays contained internal positive controls to detect RT PCR inhibitors N gene real time RT PCR assay The N gene was amplified and quantified in a 25 H9262l reaction containing 5 H9262l of RNA 12 5 H9262l of reaction buffer and 0 6 H9262lof reverse transcriptase Taq mixture Superscript II RT Platinum One Step RT PCR kit Invitrogen 3 6 mM magnesium sulfate 1 H9262g of bovine serum albumin Sigma 200 nM concentrations each of primer SANS1 TGGACCCACAGA TTCAACTGA and probe SANP1 6 carboxyfluorescein TAACCAGAATGG AGGACGCAATGG 6 carboxy N N NH11032 NH11032 tetramethylrhodamine and a 400 nM concentration of primer SANPAs2 GCTGTGAACCAAGACGCAGTAT Thermal cycling involved 50 C for 10 min followed by 95 C for 3 min and then 50 cycles of 95 Cfor2s 55 C for 12 s and 72 C for 10 s Fluorescence F1 F2 was read at 55 C on the LightCycler Quantification of RNA was done by using cloned and in vitro transcribed RNA standards as described earlier 2 3 Inhi bition control was performed by parallel testing of a second aliquot of each sample spiked with in vitro transcribed RNA standard in a concentration 10 fold above the sensitivity limit of the assay No cross reaction of the test was observed with human coronaviruses 229E and OC43 as well as with various animal coronaviruses No positive results were obtained in tests of 54 stool samples from diarrheic patients and 30 respiratory samples sputum and throat swab speci mens from non SARS patients with respiratory disease symptoms mostly atyp ical pneumonia Virus culture Virus replication was monitored by inoculating SARS CoV onto subconfluent Vero cell cultures at a multiplicity of infection of 0 01 After 1 to 4 days of incubation at 37 C and 5 CO 2 the supernatant was removed and cleared by centrifugation at 10 000 H11003 g for 10 min Determination of analytical sensitivities of RT PCR methods To make sure that the analytical sensitivity of the in house N gene assay was generally equiv alent to that of commercial test kits its lower detection limit was compared to that of one of the R gene assays RealArt HPA coronavirus LC kit by probit analysis as previously described 2 with cloned and in vitro transcribed RNA calibrators A 95 detection chance was achieved with 2 8 or 3 0 copies per reaction in the Artus and N assays respectively According to the Poisson distribution formula this reflects a true detectability of one copy of RNA per reaction in both tests i e P a H11005 e H11002m m a a where P is the probability of a positive occurrences in an 100 efficient test at an average of m objects per volume unit set P H11005 0 05 a H11005 0 Statistical analysis All statistical calculations were done with the Statgraph icsPlus software package version 5 0 Statistical Graphics Corp RESULTS All 66 samples from 29 confirmed SARS patients were tested with the RealArt HPA coronavirus LC kit and the LightCycler SARS CoV quantification kit Quantitative results were recorded for all samples that tested positive Samples with negative test results for SARS CoV but no amplification of the internal positive control were excluded from the evalu ation due to PCR inhibition Inhibition occurred in one sample in the Artus assay and in two samples in the Roche assay All samples were also tested in an N gene real time RT PCR assay to determine whether detecting this gene of SARS CoV might further increase the sensitivity of RT PCR The data were recorded as described above No RT PCR inhibition occurred with this test Table 1 summarizes the qualitative test results by sample type Virus RNA was detected in only ca 70 of samples depending on the test used No significant difference in sensi tivity between the individual assays could be observed as de termined by an analysis of variance P H11005 0 680 F ratio H11005 0 39 Although all three assays detected SARS CoV in all 100 samples from the lower respiratory tract only 56 to 58 of upper respiratory tract samples and 78 to 87 of stool samples tested positive When all three RT PCR test results per sample were taken into account 71 2 of samples were positive in at least two of three tests and 80 3 of samples were positive in at least one of three tests In the 11 patients from whom more than one sample was available three to seven samples per patient median four samples each assay detected the virus in at least one sample per patient Figure 1 shows that the results of RNA quantification in both commercial R gene kits correlated well with those of the N gene test There was no significant difference of N and R gene copy numbers in the samples in a Wilcoxon matched pairs signed rank test log 10 copies in the N assay versus log 10 copies in the Artus kit P H11005 0 5865 N assay versus the Roche kit P H11005 0 3897 As depicted in Fig 2 none of the three main categories of samples commonly taken from SARS patients specimens from the deep respiratory tract the upper respira tory tract and stool yielded significantly more N gene than R gene RNA The median concentrations of R and N gene RNA respectively were 5 5 H11003 10 2 and 5 2 H11003 10 2 copies sample in throat swabs and saliva 1 2 H11003 10 6 and 2 8 H11003 10 6 copies ml in sputum and endotracheal aspirates and 4 3 H11003 10 4 and 5 5 H11003 10 4 copies ml in stool This suggests that the N assay also would not yield a better sensitivity with a particular sample type TABLE 1 Clinical sensitivities of three different SARS CoV RT PCR assays Sample origin a Median day of disease range b Sensitivity c in confirmed patients as determined by various methods target gene Artus kit R Roche kit R In house N Upper respiratory tract d 10 5 3 25 11 19 10 18 g 11 19 Lower respiratory tract e 12 3 34 12 12 12 12 12 12 Stool 12 5 8 37 18 22 g 18 23 20 23 Plasma 9 3 14 1 7 2 7 3 7 Other f 31 2 54 4 5 1 4 3 5 a n H11005 66 samples The samples were from 29 laboratory confirmed SARS patients 25 patients with seroconversion in an immunofluorescence assay and 4 patients with a positive RT PCR result in three independent assays equivalent to World Health Organization criteria for laboratory confirmation of SARS b That is the day at which the sample was obtained c Expressed as the number of positive samples total number of samples The total percent sensitivities 95 confidence intervals for the Artus Roche and in house methods were 70 8 59 to 82 67 1 55 to 79 and 74 2 63 to 85 respectively d Three saliva samples and sixteen nasopharyngeal swabs e Eight sputum samples two endotracheal aspirate specimens and two bron choalveolar lavage samples f One jejunectomy sample one dialysis fluid specimen and three urine sam ples g There was no result for one sample due to a failed internal control The sample was therefore omitted from the sensitivity evaluation for this assay 2044 DROSTEN ET AL J CLIN MICROBIOL on June 17 2015 by UPVA http jcm asm org Downloaded from RNA concentrations between the categories of samples dif fered significantly for both RNA species measured Wilcoxon two sample tests comparing the virus concentrations in the whole cohort by category of sample yielded P values from 0 001 to H113490 00001 Thus successful detection of SARS CoV de pended more on the type of sample tested than on the type of test used note that the time of sampling was equivalent in all categories of samples As shown previously samples from the lower respiratory tract contained the highest concentration of RNA 3 From four patients samples from the upper and lower respiratory tract were obtained on the same day The virus RNA concentrations in the lower respiratory tract were clearly higher than in the upper respiratory tract in each pa tient means of three quantitative tests per sample in the four patients the lower respiratory tract samples contained 9 H11003 10 5 6 H11003 10 7 3 2 H11003 10 5 and 1 8 H11003 10 7 more RNA copies ml than the respective upper respiratory tract samples Virus concentrations in plasma and urine were very low with either assay ranges 7 2 H11003 10 1 to 5 6 H11003 10 3 copies ml plasma and 1 5 H11003 10 2 to 4 7 H11003 10 4 copies ml urine To appreciate why the previously reported higher intracel lular abundance of the N gene was not reflected in clinical diagnostic results we quantified SARS CoV N and R gene RNA in Vero cell cultures 1 to 4 days after infection Fig 3 The cytoplasm of the cells yielded about five times more N than R RNA after 1 day On subsequent days however levels of both RNAs gradually converged and approximated each other In the supernatant analyzed as a control the abundance of both RNAs was always equivalent DISCUSSION In the present study we have determined whether the clinical detection of SARS CoV RNA can be improved by novel com mercial RT PCR kits or by using an alternative target gene for detection N gene instead of R gene The rationale for using commercial tests was that these are highly optimized and con tain improved technical features e g internal controls Both kits detected the virus in ca 70 of all samples which appears to be better than what has been observed in an earlier study 7 The sensitivity in stool samples reached 78 to 87 versus FIG 1 Linear regression of the virus RNA concentrations deter mined with the N assay y axis and each of the two commercial R assays x axis suppliers of R assays are identified in the panels in clinical specimens of SARS patients from Singapore Hong Kong and Germany that yielded positive results in both assays r H11005 correlation coefficient For swab samples virus concentrations are expressed in terms of copies per swab for all other samples the virus concentration is given as copies per milliliter of sample FIG 2 Box plot analysis of concentrations of R and N gene RNA y axis in three different categories of clinical samples x axis ETA H11005 endotracheal aspirate The upper and lower limits of the boxes rep resent the innermost two quartiles of the ranked datasets whereas the lines represent the outermost quartiles Horizontal lines within boxes represent the medians crosses depict the means of the datasets The P values associated with each category of samples are derived from Wilcoxon matched pair signed rank tests comparing the observed con centrations of the two different RNA species in each category of samples VOL 42 2004 SARS RT PCR 2045 on June 17 2015 by UPVA http jcm asm org Downloaded from 58 to 63 with earlier methods 10 On the other hand the sensitivity in samples from the upper respiratory tract was not clearly improved with the refined commercial tests and the new tests still do not appear to allow ruling out SARS reliably in suspected patients Tests for ruling out an infection with SARS CoV are ur gently required because the disease is highly contagious and cannot be distinguished by other means from common respi ratory diseases e g influenza We thus attempted to improve the sensitivity further by switching the RT PCR target to the N gene based on the reportedly higher abundance of such sub genomic RNA in infected cells 9 Use of this gene for diag nostic purposes has been encouraged by first animal model data suggesting that it might improve the detectability of the virus 6 In our patients however this approach did not yield the expected benefit even though our N gene assay was capa ble of detecting one RNA molecule per reaction a sensitivity that cannot be improved further Our cell culture data give an explanation for this surprising finding Although it could be confirmed that the N gene is more abundant than the R gene in freshly infected cells the excess decreased after a relatively short time Since it can be assumed that by the time symptoms appear cells in most clin ical specimens would have already been replicating the virus for some days this explains why the sensitivity of the N assay is not superior in symptomatic patients When we analyzed the abundance of the N gene by sample type this assumption was confirmed No category of samples specimens from the upper respiratory tract lower respiratory tract and stool contained more N gene than R gene RNA Apart from the target gene used time and body compart ment of sampling are other possible factors influencing the sensitivity of RT PCR The timing of samples in our study was comparable to that in earlier reports 7 10 representing the acute febrile phase of disease when patients are hospitalized and the probability of detection of RNA is high especially in respiratory samples 7 The sensitivity of RT PCR in our SARS patients was significantly influenced by the body com partment a sample was taken from nasopharyngeal swabs and saliva yielded significantly less viral RNA than other samples Although they can be easily taken from patients swabs and saliva thus have to be considered suboptimal material for SARS CoV RT PCR Sputum and endotracheal aspirates can be expected to give better diagnostic sensitivity but their col lection poses the risk of generating infectious aerosols in the hospital Stool samples appear to be a reasonable alternative