DISCUSSION
This analysis compared the infectivity and virulence of two SMV inocula used in two different human challenge trials by examining infection rates, illness rates, severity of illness, viral shedding, and IgG seroconversion among subjects challenged with a single dose of one of these inocula. Inoculum 1 was prepared at NIH sometime between 1977 and 1979 from the stool filtrate of a subject infected in the original Snow Mountain Virus outbreak in Colorado in December 1976. This safety-tested inoculum was first used in a human challenge study with 12 subjects conducted at NIH by Dolin et al. [12] and then again in a human challenge study with 15 subjects conducted between 2000-2002 at the University of North Carolina-Chapel Hill [11]. Inoculum 2 was prepared from the stool filtrate of an infected subject in the 2000-2002 human challenge trial and was then used in a human challenge study with 44 subjects at Emory University from 2016-2018 [9]. The goal of this analysis is to examine whether the infectivity, pathogenicity, and virulence of the SMV changed due to passage in the human host by assessing differences in the outcome measures of the two more recent challenge trials.
Both inocula were associated with similar overall infection rates as measured by the proportion of challenged subjects who developed infection and illness after challenge in each trial. However, the median infectious dose (ID50) of Inoculum 2 was about 100-fold higher than that estimated for Inoculum 1 [13]. Comparing the severity scores of infected subjects with clinical symptoms indicated that subjects infected with SMV Inoculum 2 had less severe acute gastroenteritis and more delayed viral shedding even though these subjects were challenged with a much higher dose than the subjects who were infected with Inoculum 1. These results are consistent with what we recently reported [15] that subjects challenged with a first-generation NV inoculum (8FIIa, prepared in 1971) had significantly higher severity scores of acute gastroenteritis but shorter duration of viral shedding compared with those challenged with a second-generation NV inoculum (8FIIb, prepared in 1997 from the stool filtrate of a subject infected with the first NV inoculum).
Viral shedding and clinical illness are two important outcomes evaluated in NoV challenge trials. Based on outbreak investigations and human challenge studies, most infected subjects experience clinical symptoms along with viral shedding in their feces during the first several days of infection [16]. Symptoms usually resolve after 48-72 hours, however viral shedding can continue for up to three weeks [16] and even longer in immunocompromised subjects [17, 18]. In this study, all subjects infected with SMV Inoculum 1 shed virus for up to six days, but 3% of subjects infected with Inoculum 2 had viral shedding between days 15 to 45 days post challenge. We hypothesize that the less severe clinical illness, but possibly longer SMV shedding, observed in subjects infected with Inoculum 2 may be associated with viral mutations that reduced the virulence of the inoculum, rather than host factors or laboratory assessment methods. However, it is possible that the RT-qPCR assay used to measure viral shedding in the more recent challenge trial with Inoculum 2 is more sensitive and better able to detect prolonged shedding compared to the conventional RT-PCR assay used in the trial with Inoculum 1 [19].
As the prototype of NoV GII.2, SMV is associated with a small percentage of NoV outbreaks, and little is known about the mutation of this virus. Swanstrom et al., [20] reported that the sequence of the P2 domain of SMV strains collected between 1976 and 2010 has evolved over time but less extensively than has been reported for GII.4 NoV strains. Mutations in the surface protein could significantly alter the antigenicity of representative strains, which could also change viral function, pathogenesis, transmission, and infectivity. Studies of other RNA viruses indicate that viral mutations can be deleterious or favorable to the pathogens in terms of infectivity and virulence. Both host and viral factors may explain the change in infectivity and virulence that we observed between the two SMV inocula. Some host factors, including acquired and innate immunity prior to challenge and secretor status, can affect the likelihood of NoV infection and illness. Other host factors, including age, sex, and race/ethnicity may or may not impact the risk of NoV infection and illness. Rouphael et al [9] reported that prechallenge anti-SMV serum IgG concentration, carbohydrate-binding blockade antibody, and salivary IgA were not associated with infection with Inoculum 2. Given the evidence from viral evolution of other RNA viruses such as influenza and SARS-CoV-2 [21-23], we hypothesize that the less severe symptoms associated with the second generation SMV inoculum in this study may be due to intra-host SMV mutations during the course of infection in the subject who was the original source of Inoculum 2 and possibly further inter-host mutations in the subjects who were infected with Inoculum 2 after challenge.
This analysis is the first to compare the clinical outcomes associated with two different SMV inocula used in two human challenge trials [9,11]. We examined detailed clinical and laboratory data collected from well-controlled human challenge studies with nearly identical protocols that included follow-up sample collection and analyses to carefully characterize infection, illness, and immune response. The limitations of this analysis include small sample size in both studies that may not provide enough power to detect a significant difference in some outcomes, retrospective analysis of two studies that span over 20 years and had different study populations, and slightly different laboratory methods for measuring viral shedding and serum antibodies. Finally, we were not able to assess how differences between inocula preparation in the late 1970’s (Inoculum 1) vs. 1997 (Inoculum 2) and storage may have affected the infectivity, pathogenicity, and virulence of these two inocula. The time between inoculum preparation and the challenge trials that provided the data we analyzed was approximately 20 years for both inocula: Inoculum 1 (1979-2000) and Inoculum 2 (1997-2016).
Understanding the difference in the severity of illness associated with these two SMV inocula is critical for those who plan to use second generation NoV inoculum to evaluate new NoV vaccine candidates in vaccination-challenge trials. As we observed for the NV inocula, the second generation SMV Inoculum 2 was associated with a higher ID50 and less severe clinical illness when compared with the first generation SMV inoculum. If second generation NoV inoculum does not elicit expected symptomatic illness rates in challenged subjects, it will become more complex and costly to conduct vaccination-challenge trials because they may require larger numbers of study subjects and challenges with higher doses of inoculum in order to assess the efficacy of a candidate vaccine to reduce illness.
Funding: This study was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health (UL1TR002378); the National Institute of Allergy and Infectious Disease (R01 AI148260); and the Division of Microbiology and Infectious Diseases to the Emory Vaccine and Treatment Evaluation Units (VTEU): HHSN272200800005C, HHSN272201300018I, HHSN27200003, and HHSN27200018
Table 1: Modified Vesikari score (18-point scale) [14] for SMV-infected subjects
with clinical symptoms