key: cord-0749731-mpq3oxup authors: Tian, Di; Lin, Zhen; Kriner, Ellie M.; Esneault, Dalton J.; Tran, Jonathan; DeVoto, Julia C.; Okami, Naima; Greenberg, Rachel; Yanofsky, Sarah; Ratnayaka, Swarnamala; Tran, Nicholas; Livaccari, Maeghan; Lampp, Marla; Wang, Noel; Tim, Scott; Norton, Patrick; Scott, John; Hu, Tony Y.; Garry, Robert; Hamm, Lee; Delafontaine, Patrice; Yin, Xiao-Ming title: Ct Values do not Predict SARS-CoV-2 Transmissibility in College Students date: 2021-06-05 journal: J Mol Diagn DOI: 10.1016/j.jmoldx.2021.05.012 sha: f0bd4d80518752e11f4d176e6610eabe09fbfa0c doc_id: 749731 cord_uid: mpq3oxup SARS-CoV-2 is highly contagious, and the global spread has caused significant medical/socioeconomic impacts. Other than vaccination, effective public health measures, including contact tracing, isolation and quarantine, is critical for deterring viral transmission, preventing infection progression and resuming normal activities. Viral transmission is affected by many factors but the viral load and vitality could be among the most important ones. Although in vitro studies have indicated that the amount of virus isolated from infected people affects the successful rate of virus isolation, whether the viral load carried at the individual level would determine the transmissibility was unknown. From the diagnostic point of view, we aimed to examine whether the Ct value, a measurement of viral load by RT-PCR assay, could differentiate the spreaders from the non-spreaders in a population of college students. Our results indicate that while at the population level the Ct value is lower, suggesting a higher viral load, in the symptomatic spreaders than that in the asymptomatic non-spreaders, there is significant overlap in the Ct values between the two groups. Thus Ct value, or the viral load, at the individual level could not predict the transmissibility. Our studies also suggest that a sensitive method to detect the presence of virus is needed to identify asymptomatic persons who may carry a low viral load but can still be infectious. The rapid spread of SARS-CoV-2 has caused a global pandemic with serious impact on all aspects of human life. Deterrence of viral transmission through public health measures, including contact tracing, isolation and quarantine, is critical for infection control required to resume normal activities. Unlike two other betacoronavirus that had caused previous local epidemics, SARS-CoV and MERS-CoV, SARS-CoV-2 exhibits a distinct replication and transmission kinetics. It replicates more rapidly in the human upper respiratory tract, which helps its transmission through asymptomatic viral carriers and facilitates a fast spread of SARS-CoV-2. The relatively lower fatality rate (CFR or case fatality ratio) of SARS-CoV-2 (2% compared to SARS-CoV's 10% and MERS-CoV's 34%) may also contribute to its high transmissibility 1 . SARS-CoV-2 is highly contagious with an estimated reproductive number (Ro) of 3.5 2 , but significant variations exist among individuals with some being super spreaders. This is much higher than the Ro of seasonal flu (1.3) and SARS-CoV (0.86-1.83) 3, 4 . The viral load in an infected person could affect the level of infectivity. Several studies have found that successful isolation of virus from patient samples depended on viral load as measured by the cycle threshold (Ct) value of the RT-PCR assay, which was thus suggested to correlate with infectivity [5] [6] [7] [8] [9] [10] . A cutoff Ct value between 32 and 35 was proposed to guide isolation practices [5] [6] [7] [8] [9] [10] . However, it was not clear whether the in vitro culture results could reflect actual viral spread in persons and whether Ct values could actually be used to guide isolation and quarantine decisions. The effective way to block the viral transmission is to identify, isolate and treat the infected persons, and to track down and quarantine those having close contact with the infected ones. As the infection involves more and more people, individual communities or regions may be forced to be shut down. All social activities related to work, study and leisure will be significantly affected with tremendous impacts on the economy, the society and the overall personal health condition. It is thus important to understand better the dynamics of viral transmission and examine whether certain surrogate measurement may be used to determine SARS-Cov-2 transmissibility. We thus aimed to determine whether the Ct values, as a measurement of viral load, could be used to provide a level of prediction in a population of college students. We J o u r n a l P r e -p r o o f compared the Ct values of the spreaders and that of the non-spreaders, and found these values were largely overlapping. It is thus not possible to predict viral transmissibility based on Ct values at the individual level. J o u r n a l P r e -p r o o f Study population. Results from undergraduate students at the age of < 23 years of old were selected for this retrospective study. These students were participants in the on-campus education activities while living either on campus or off campus. They were tested at the frequency of twice a week in the period between September 1 st 2020 and October 31 st , 2020. This study included only students who were tested in the CLIA-certified Molecular Pathology Laboratory of the Department of Pathology and Laboratory, Tulane University School of Medicine because the Ct values were obtained using the same testing method in the same laboratory for all the included subjects. Full review and approval is waived by the Tulane University IRB due to involvement of only secondary, deidentified data. Sample collection, processing and RNA extraction. Nasopharyngeal swab specimens were collected following current CDC guidelines. All samples were stored at 4°C before delivering to the testing laboratory. Upon receiving, samples were inactivated at 60°C for 30min in a forced- Contact tracing and quarantine program. Symptomatic information was collected immediately before sample collection and testing. Contact Tracers received all positive results and made phone calls to reach positive cases. They interviewed the positive cases to identify close contacts. In addition, they helped to establish the quarantine procedure. The information of the index cases and the contacted was recorded. Statistical analysis. Data are presented as mean±SD (for the age distribution), mean ± SEM, or median ± interquartile (for the Ct values). The statistical significance is assessed by two-sided unpaired t test for age distribution, Mann-Whitney U test or one-way ANOVA for Ct values using Prism Software (version 9, Graphpad Software, Inc. San Diego, CA). Colleges represent a unique environment with a dense population of primarily young students and strict control of SARS-CoV-2 transmission is critical for their education mission. Tulane University maintained on-campus educational activities in the fall semester of 2020. We established a high throughput SARS-CoV-2 testing program to support the contact tracing, isolation and quarantine efforts needed to actively restrict viral transmission throughout the campus. During the period covered in this study, we performed the screening test at the frequency of twice a week with 99% of testing completed within 24 hours from collection to report. While all students (graduate and undergraduate, on-campus and off-campus living) were screened, only data from 7,440 students under the age of 23 years old from September 1 st 2020 to October 31 st , 2020 were included in this study for data consistency. We performed a total of 61,982 tests for these students during this period and identified 602 unique positive cases (Tables 1-2) . Compared to all the students, those tested positive for SARS-CoV-2 were slightly younger, reflecting that more freshmen and sophomores were infected. In addition, male and female students had nearly the same proportion of the infected (49.3% vs 50.7%), consistent with a meta-analysis of 90 reports 11 . However, considering that male students accounted for only 37.5% of all the students screened, the male students had a higher infection rate (10.65%) than the female students (6.56 %) in this cohort. From this cohort of 602 positive individuals, we identified 195 index cases with one or more reported close contacts who were then tested during their mandated 14-day quarantine period for the evidence of transmission from their associated index cases (Fig. 1A) . We found that 48.2% (94/195) of these index cases had at least one contact who became SARS-CoV-2-positive, whereas 51.8% of the index cases (n=101) were non-spreaders with no contacts who subsequently tested positive. Mean Ct values of the spreaders and the non-spreaders were nearly identical (Fig. 1B) , but their median Ct values differed by almost one cycle (Fig. 1C) , suggesting that more spreaders had a lower Ct value than the non-spreaders. However, Ct distributions in these groups were similar with the main peaks around 18-21 (Fig. 1D) , although the Ct range was slightly broader for the spreaders (12-36) than that for the non-spreaders (14-36). Cumulative Ct frequencies overlapped between the spreaders and the non-spreaders with 10.9%, and 13.8% of cases having J o u r n a l P r e -p r o o f a Ct value of 32 and higher, respectively (Fig. 1E ), but the difference was not large enough to discriminate the two groups for practical use. In a reverse approach, index cases were traced for 481 students undergoing quarantine at one of the three Tulane quarantine sites in September 2020 ( Fig. 2A) , 18% of whom (85/481) became positive during their quarantine period. Index cases for these 481 quarantined individuals were considered spreaders if they were linked to one or more quarantined students with a positive test result, or non-spreaders if they were associated only with individuals with negative test results. Spreaders and non-spreaders without Ct reported were excluded from further analysis. We found that mean Ct values of the spreader and the non-spreader groups did not differ (Fig. 2B) . Taken together, these index case studies suggest that Ct values alone do not predict transmission risk. Individuals who are SARS-CoV-2 positive but asymptomatic can still be infectious [12] [13] [14] , and may exhibit a similar viral load to their symptomatic counterparts 12, 13, 15 . We therefore identified 375 positive cases who were evaluated for COVID-19 symptoms at testing to assess the relationship between symptom presentation and Ct values (Fig. 3A) . The reported symptoms included lethargy, fever, headache, cough, running nose, and gastrointestinal symptoms. We found that the mean and median Ct values were significantly lower in symptomatic than those in asymptomatic cases (Fig. 3B-C) , which was also reflected by the difference in the Ct range of these groups (12-36 versus 14-37; Fig. 3D ). Although both groups exhibited Ct peaks around 19-22, there was a noticeable rightward shift in the cumulative Ct frequency in the asymptomatic versus symptomatic population, indicative of reduced viral load in the asymptomatic group (Fig. 3E ). In comparison, other studies with cohorts differing in location and in constituents, including a large study involving senior citizens from nursing houses and assisted living facilities in Massachusetts, found that Ct values did not differ significantly between the symptomatic and the asymptomatic individuals; but observed a faster virus clearance, as measured by Ct value, in the asymptomatic cases than in the symptomatic cases 13, 15 . These and our studies thus suggest that infections with a higher viral load may more likely lead to symptom development, or that symptomatic persons tend to have higher viral loads or to maintain their viral loads for a longer time. All 195 index cases with contact tracing information had information recorded regarding symptoms. We thus further divided the spread group and the non-spreader group based on symptom presentation (Fig. 4A) . We found that the symptomatic spreaders had the lowest mean and median Ct values, differing by 2 cycles for the mean and 3.5 cycle for the median when compared with the asymptomatic non-spreaders, which had the highest mean and median Ct values ( Fig. 4B-C) . The Ct distribution indicated that the symptomatic groups (spreaders and non-spreaders) and the spreader groups (with or without symptoms) tended to have more individuals with lower Ct values (<24) (Fig. 4D-E) . This finding suggests that SARS-CoV-2 spreaders tend to have higher viral loads and are more likely symptomatic. The present study for the first time compared the Ct values between the spreaders and the non-spreaders of SARS-CoV-2 infected persons in a college student population. We have found that while the mean Ct values of the spreaders, particularly the symptomatic spreaders, are lower than the non-spreaders, there are significant overlaps among individuals, whether they are spreaders or non-spreaders. It is thus practically not feasible to predict who would be spreaders based on the viral load as detected from the nasal swab. Ct values are not reported in current public health practice despite that they may be informative of viral burden. Our study supports this practice and indicates that, due to the broad spread and overlap in Ct values across the spectrum of symptom presentation and transmissibility, Ct value reporting at the individual level, such as by setting a cut-off value at 32 5-10 , would provide little diagnostic value for differential case management. At the population level, Ct values may be useful, particularly in association with the symptomatic presentation, to indicate the likelihood of transmission. It may thus have epidemiological or surveillance values. Detection of SARS-CoV-2 may need to be both sensitive and rapid, which may not always be achieved by all methods. Rapid but less sensitive method should be used more frequently in order to catch individuals whose virus level may be elevating over the course of infection and thus presumably become more infectious. However, our results suggest that individual with low viral load could still be infectious. Thus, a sensitive and robust SARS-CoV-2 diagnostic testing method is needed to effectively control the viral transmission by maximizing the ability to identify and quarantine those infected with a low level of virus. Although limited by its retrospective nature, this study has the advantage of being less interfered by the host and environmental factors on viral transmission, since the college student population is generally in good health with few underlying susceptibilities, with most individuals living and interacting in a shared and relatively confined social environmental (i.e., campus). We have further restricted data to those from current undergraduate students (<23 years old, with an average age of 20.3 years old ( Table 2) ), making the population more homogenous to reduce the influence of the age. Transmissibility is not only affected by the viral load of the spreaders and the environment where transmission takes place, but also affected by factors that underline the susceptibility of the population, such as the age, sex, and the basic health conditions. From J o u r n a l P r e -p r o o f this aspect, it is interesting to note that while male and female students had nearly the same proportion of the infected, consistent with a meta-analysis of 90 reports 11 , the male students did have a higher infection rate (10.65%) than the female students (6.56 %) in this cohort. The sex disparity of COVID-19 has been well recognized in terms of the severity of the disease with the male being more likely to develop severe conditions 11 . It has yet to be determined how the sex makes the difference in the COVID-19 spread and development. In summary, this study has determined that Ct values of the spreaders may be lower at the population level than the non-spreaders, but the large overlap in the values at the individual level prevents their use as a differential diagnostic tool to guide isolation and quarantine practice. On the other hand, a sensitive and robust diagnostic method is necessary to restrict viral transmission from those carrying a low level of virus. J o u r n a l P r e -p r o o f Cases 36 34 32 30 28 26 24 38 20 18 16 14 12 22 36 34 32 30 28 26 24 38 20 18 16 14 12 22 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Cases 34 32 30 28 26 24 20 18 16 14 12 22 36 34 32 30 28 26 24 38 20 18 16 14 12 22 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 32 30 28 26 24 38 20 18 16 14 12 22 36 34 32 30 28 26 24 38 20 18 16 14 12 22 36 34 32 30 28 26 24 38 20 18 16 14 12 22 34 32 30 28 26 24 20 18 16 14 Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72314 Cases From the Chinese Center for Disease Control and Prevention Estimating the reproductive number and the outbreak size of COVID-19 in Korea Estimates of the reproduction number for seasonal, pandemic, and zoonotic influenza: a systematic review of the literature Model parameters and outbreak control for SARS Viral cultures for COVID-19 infectivity assessment -a systemic review Viral RNA load as determined by cell culture as a management tool for discharge of SARS-CoV-2 patients from infectious disease wards La Scola B: Correlation between 3790 qPCR positives samples and positive cell cultures including 1941 SARS-CoV-2 isolates Predicting infectious SARS-CoV-2 from diagnostic samples Duration of infectiousness and correlation with RT-PCR cycle threshold values in cases of COVID-19 Cell-based cutlure of SARS-CoV-2 informs infectivity and safe de-isolation assessments during COVID-19 Male sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU admission Presumed Asymptomatic Carrier Transmission of COVID-19 group OC-r: The natural history and transmission potential of asymptomatic SARS-CoV-2 infection Temporal dynamics in viral shedding and transmissibility of COVID-19 Comparision of viral levels in individuals with or without symptoms at time of COVID-19 testing among 32,480 residents and staff of nursing homes and assisted living facilities in Massachusetts percentage of each population at the Ct value of 24. At this Ct value and below, there is a higher percentage of symptomatic spreader cases (66.2%) than asymptomatic non-spreader cases (48.7%). The percentage of cases of the other groups are between the two. J o u r n a l P r e -p r o o f