Healthcare costs for hospitalized COVID-19 patients in a Japanese university hospital: A cross-sectional study (preprint)

Background: A health-economic evaluation related to COVID-19 is urgently needed to allocate healthcare resources efficiently; however, relevant medical cost data in Japan concerning COVID-19 are scarce. Methods: This cross-sectional study investigated the healthcare cost for hospitalized COVID-19 patients in 2021 at Keio University Hospital. We calculated the healthcare costs during hospitalization using hospital claims data and investigated the variables significantly related to the healthcare cost with multivariate analysis. Results: The median healthcare cost per patient for the analyzed 330 patients was Japanese yen (JPY) 1,304,431 (US dollars ~11,871) (interquartile range: JPY 968,349–1,954,093), and the median length of stay was 10 days. The median healthcare cost was JPY 798,810 for mild cases; JPY 1,113,680 for moderate I cases; JPY 1,643,909 for moderate II cases; and JPY 6,210,607 for severe cases. Healthcare costs increased by 4.0% for each additional day of hospitalization; 1.26 times for moderate I cases; 1.64 times for moderate II cases; 1.84 times for severe cases; and 2.05 times for cases involving ICU stay. Conclusions: We clarified the healthcare cost for hospitalized COVID-19 patients by severity in a Japanese university hospital. These costs contribute as inputs for forthcoming health economic evaluations for strategies for preventing and treating COVID-19.

Characteristics of patients with COVID-19 who have deteriorating chest X-ray findings within 48 hours: a retrospective cohort study (preprint)

Background The severity of chest X-ray (CXR) findings is a prognostic factor in patients with coronavirus disease 2019 (COVID-19). However, the prognostic impact of deterioration of CXR findings and the clinical characteristics of patients with worsening CXR findings remain unclear. We aimed to investigate the clinical and genetic characteristics, as well as the prognosis, of patients with worsening CXR findings during early hospitalisation.Methods We retrospectively included 1656 consecutive Japanese patients with COVID-19 recruited through the Japan COVID-19 Task Force. Rapid deterioration of CXR findings was defined as increased pulmonary infiltrates in ≥ 50% of the lung fields within 48 h of admission.Results Rapid deterioration of CXR findings was an independent risk factor for death, most severe illness, tracheal intubation, and intensive care unit admission. The presence of consolidation on CXR, comorbid cardiovascular and chronic obstructive pulmonary diseases; high body temperature (≥ 37.7°C); and increased levels of serum aspartate aminotransferase (≥ 30 IU/L), potassium (≥ 4.3 mEq/L), and C-reactive protein (≥ 2.53 mg/dL) were independent risk factors for rapid deterioration of CXR findings. The risk variant at the ABO locus (rs529565-C) was associated with rapid deterioration of CXR findings in all patients with COVID-19. Further, the population-specific risk variant at the DOCK2 locus (rs60200309-A) was nominally associated with rapid deterioration of CXR findings in patients aged < 65 years.Conclusions This study revealed the clinical features, genetic features, and risk factors for rapid deterioration of CXR findings in patients with COVID-19. Rapid deterioration of CXR findings is a poor prognostic factor for patients with COVID-19.

T-cell Response and Antibody Production by Booster COVID-19 Vaccination in Japanese Patients with Chronic Kidney Disease Treated with Hemodialysis (preprint)

Most studies on vaccines of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have focused on antibody, but cellular immunities are also critical. We aimed to evaluate the immune reactions of hemodialysis (HD) patients after the administration of the booster dose from the perspective of both humoral and cellular immunities. Hemodialysis patients (HD group) and age- and sex-matched non-dialysis individuals (control group) receiving three doses of BNT162b2 vaccine were measured for anti-SARS-CoV-2 immunoglobulin (IgG) and T-SPOTⓇ.COVID test (T-SPOT) before, 3 weeks, and 3 months after the booster dose. The HD group had significantly higher SARS-CoV-2 IgG levels 3 weeks and 3 months after the booster dose than the control group, although both groups had no difference in SARS-CoV-2 IgG levels before the booster dose. Moreover, the HD group had significantly higher T-SPOT levels before and 3 weeks after the booster dose than the control group, but the difference was not significantly different 3 months after the booster dose. Furthermore, the incidence rates of local and systemic adverse reactions were significantly higher in the HD group than in the control group. HD patients obtained higher SARS-CoV-2 IgG levels and SARS-COV-2-specific T-cell responses after the booster dose than control.

Memory B cells and memory T cells induced by SARS-CoV-2 booster vaccination or infection show different dynamics and efficacy to the Omicron variant. (preprint)

Although BNT162b2 vaccination was shown to prevent infection and reduce COVID-19 severity, and the persistence of immunological memory generated by the vaccination has not been well elucidated. We evaluated memory B and T cell responses to the SARS-CoV-2 spike protein before and after the third BNT162b2 booster. Although the antibody titer against the spike receptor-binding domain (RBD) decreased significantly 8 months after the second vaccination, the number of memory B cells continued to increase, while the number of memory T cells decreased slowly. Memory B and T cells from unvaccinated infected patients showed similar kinetics. After the third vaccination, the antibody titer increased to the level of the second vaccination, and memory B cells increased at significantly higher levels before the booster, while memory T cells recovered close to the second vaccination levels. In memory T cells, the frequency of CXCR5+CXCR3+CCR6- cTfh1 was positively correlated with RBD-specific antibody-secreting B cells. Furthermore, T cell-dependent antibody production from reactivated memory B cells in vitro was correlated to the Tfh-like cytokine levels. For the response to variant RBDs, although 60%-80% of memory B cells could bind to the Omicron RBD, their binding affinity was low, while memory T cells show an equal response to the Omicron spike. Thus, the persistent presence of memory B and T cells will quickly upregulate antibody production and T cell responses after Omicron strain infection, which prevents severe illness and death due to COVID-19.

Age stratified seroprevalence of antibodies against SARS CoV 2 in the pre and post vaccination era, February 2020 to March 2022, Japan (preprint)

Japan has reported a small number of COVID-19 cases relative to other countries. Because not all infected people receive diagnostic tests for COVID-19, the reported number of COVID-19 cases must be lower than the actual number of infections. Assessments of the presence of antibodies against the spike protein of SARS-CoV-2 can retrospectively determine the history of natural infection and vaccination. In this study, we assessed SARS-CoV-2 seroprevalence by analyzing over 60,000 samples collected in Japan from February 2020 to March 2022. The results showed that about 5% of the Japanese population had been infected with the virus by January 2021. The seroprevalence increased with the administration of vaccinations to adults; however, among the elderly, it was not as high as the vaccination rate, probably due to poor immune responses to the vaccines and waning immunity. The infection was spread during the epidemic waves caused by the SARS-CoV-2 Delta and Omicron variants among children who were not eligible for vaccination. Nevertheless, their seroprevalence was as low as 10% as of March 2022. Our study underscores the low incidence of SARS-CoV-2 infection in Japan and the effects of vaccination on immunity at the population level.

Association between ABO blood group/genotype and COVID-19 in a Japanese population (preprint)

Background: An association between coronavirus disease 2019 (COVID-19) and ABO blood group has been reported. However, such an association has not been studied in the Japanese population on a large scale. Additionally, little is known about the association between COVID-19 and ABO genotype. Therefore, the present study aimed to investigate the association between COVID-19 and ABO blood group/genotype in a large Japanese population. Methods: All Japanese patients diagnosed with COVID-19 were recruited through the Japan COVID-19 Task Force between February 2020 and October 2021. We conducted a retrospective cohort study involving 1,790 COVID-19 Japanese patients whose DNA was used for genome-wide association study. We compared the ABO blood group/genotype in healthy population (n = 611, control) and COVID-19 patients, and then analyzed the association between the ABO blood group/genotype and clinical outcomes. Results: Blood group A was significantly more prevalent (41.6% vs. 36.8%) and group O was significantly less prevalent (26.2% vs. 30.8%) in the COVID-19 group than in the control group. Moreover, genotype OO was significantly less common in the COVID-19 group. Further, blood group AB was identified as an independent risk factor for most severe diseases compared with blood group O [aOR (95% CI) = 1.84 (1.00–3.37)]. Similarly, in ABO genotype analysis, only genotype AB was an independent risk factor for most severe disease compared with genotype OO. Conclusions: Blood group O is protective, whereas blood group A is associated with the risk of infection. Moreover, blood group AB is associated with the risk of ‘most’ severe disease. 

Estimating immunity with mathematical models for SARS-CoV-2 after COVID-19 vaccination (preprint)

Antibody waning after COVID-19 vaccinations results in the necessity of booster shots. Determining the timing of booster shots according to the decreasing antibodies of each individual will reduce the burden of excessive vaccine administration. We have developed a mathematical model to estimate waning antibody concentration (titer) during the timecourse after two doses of BNT162b2 vaccine. Using consecutive antibody datasets of over 600 vaccinated healthcare workers, we demonstrated a two-compartment model that fits better than other models. Internal validation of the model demonstrated 97.0% accuracy and external validation of healthcare worker and hemodialyzed patient datasets demonstrated 98.2% and 83.3% accuracy, respectively. A smart device application developed using our model can rapidly calculate the timing of negative seroconversion with a single antibody titer measurement. Although further modification of the model is essential, the change in antibody titer over time from COVID-19 vaccination may be estimated with a two-compartment model. 

Dynamics of antibody responses, cellular immunity, and breakthrough infections among Japanese healthcare workers during the 6 months after receiving two doses of BNT162b2 mRNA vaccine (preprint)

Introduction The waning of the antibody titre after the first two doses of the Pfizer-BioNTech BNT162b2 mRNA SARS-CoV-2 vaccine was reported. However, knowledge of the dynamics of cellular immunity is scarce. Here, we performed a prospective cohort study to disclose antibody and cellular immunity dynamics and discuss the relationship between immunity and breakthrough infection. Methods The study had a prospective cohort design. Antibody titres against SARS-CoV-2 in serially collected serum samples of 608 Japanese vaccinees after 6 months of vaccination were measured. Simultaneously, T-cell immunity dynamics were assessed using the QuantiFERON SARS-CoV-2 assay. Additionally, participants with suspected breakthrough infection were detected according to the positive conversion of the IgG assay for nucleocapsid proteins of SARS-CoV-2. Results Antibody titres were elevated 3 weeks after vaccination and waned over the remainder of the study period. The QuantiFERON SARS-CoV-2 assay performed on 536 participants demonstrated the similar dynamics. Six participants without predisposing medical conditions demonstrated positive conversion of the IgG assay for nucleocapsid proteins, while five were asymptomatic. Conclusion Waning of humoral and cellular immunity within 6 months of administration of two doses of BNT162b2 vaccine among Japanese healthcare professionals and the occurrence of asymptomatic breakthrough infection was suspected in approximately 1 of 100 vaccinees. (UMIN000043340)

Anti-SARS-CoV-2 cellular immunity in 571 vaccinees assessed using an interferon-γ release assay (preprint)

Generation of antigen-specific memory T cells has been analyzed only for few coronavirus disease 2019 (COVID-19) vaccinees, whereas antibody titers have been serologically measured for a large number of individuals. Here, we assessed the anti-severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) cellular immune response in a large cohort using interferon (IFN)-{gamma} release assays (IGRAs) based on short-term whole blood culture. The study included 571 individuals who received the viral spike (S) protein-expressing BNT162b2 mRNA SARS-CoV-2 vaccine. Serum IgG titers against the receptor-binding domain (RBD) of S protein were measured. Samples of 28 vaccinees were subjected to flow cytometry analysis of T cells derived from short-term whole blood culture. IFN-{gamma} production triggered by S antigens was observed in most individuals 8 weeks after receiving the second dose of the vaccine, indicating acquisition of T cell memory responses. The frequencies of activated T cell subsets were strongly correlated with IFN-{gamma} levels, supporting the usability of our approach. S antigen-stimulated IFN-{gamma} levels were weakly correlated with anti-RBD IgG titers and associated with pre-vaccination infection and adverse reactions after the second dose. Our approach revealed cellular immunity acquired after COVID-19 vaccination, providing insights regarding the effects and adverse reactions of vaccination.

Identification of commensals and molecular components mediating trypsin degradation in the large intestine (preprint)

The gastrointestinal tract is constitutively exposed to proteases including trypsin, a serine protease originating from the pancreas1. Elevated trypsin levels in the large intestine have been implicated in pathological conditions including infectious and inflammatory bowel disease2-4. Here we show that trypsin is regulated via degradation by members of the gut microbiota. After passing through the small intestine, trypsin activity is markedly reduced in the caecum of specific pathogen-free (SPF) mice, whereas germ-free (GF) mice have high luminal trypsin levels. We have successfully identified and isolated Paraprevotella strains from the faecal microbiome of healthy human donors as potent trypsin-degrading commensals. Mechanistically, Paraprevotella recruit trypsin to the bacterial surface through type IX secretion system-dependent polysaccharide-anchoring proteins and promote trypsin autolysis. Paraprevotella colonization protects IgA from trypsin degradation and enhances the effectiveness of oral vaccines against Citrobacter rodentium. Moreover, Paraprevotella colonization inhibits lethal infection with murine hepatitis virus, a mouse coronavirus dependent on trypsin and trypsin-like proteases for entry into host cells5,6. Congruently, carriage of putative genes involved in trypsin degradation in the gut microbiome was associated with reduced diarrhoea severity in patients with SARS-CoV2 infection. Therefore, trypsin-degrading commensal colonization may contribute to the maintenance of intestinal homeostasis and protection from pathogen infection.

Japan COVID-19 Task Force: a nation-wide consortium to elucidate host genetics of COVID-19 pandemic in Japan (preprint)

To elucidate the host genetic loci affecting severity of SARS-CoV-2 infection, or Coronavirus disease 2019 (COVID-19), is an emerging issue in the face of the current devastating pandemic. Here, we report a genome-wide association study (GWAS) of COVID-19 in a Japanese population led by the Japan COVID-19 Task Force, as one of the initial discovery GWAS studies performed on a non-European population. Enrolling a total of 2,393 cases and 3,289 controls, we not only replicated previously reported COVID-19 risk variants (e.g., LZTFL1, FOXP4, ABO, and IFNAR2), but also found a variant on 5p35 (rs60200309-A at DOCK2) that was significantly associated with severe COVID-19 in younger (<65 years of age) patients with a genome-wide significant p-value of 1.2 x 10-8 (odds ratio = 2.01, 95% confidence interval = 1.58-2.55). This risk allele was prevalent in East Asians, including Japanese (minor allele frequency [MAF] = 0.097), but rarely found in Europeans. Cross-population Mendelian randomization analysis made a causal inference of a number of complex human traits on COVID-19. In particular, obesity had a significant impact on severe COVID-19. The presence of the population-specific risk allele underscores the need of non-European studies of COVID-19 host genetics.

Pro108Ser mutant of SARS-CoV-2 3CL pro Reduces the Enzymatic Activity and Ameliorates COVID-19 Severity in Japan (preprint)

Background: It is well known that SARS-CoV-2 genome accumulates point mutations constantly. However, whether non-synonymous mutations affect COVID-19 severity through altering viral protein function remains unknown.Methods: We performed SARS-CoV-2 genome sequencing in 90 patients with COVID-19 admitted at Keio University Hospital in Tokyo Metropolitan area between March and August 2020. Viral haplotypes were examined by counting the number of non-synonymous mutations to analyse phylogenic trees and comparative amino acid sequence. Statistically relevant non-synonymous mutations were functionally evaluated with structural analyses.Findings: The number of non-synonymous mutations correlated inversely with the COVID-19 severity. Phylogenic tree analyses identified two predominant groups which were differentiated by a set of six point mutations (four non-synonymous amino acid mutations). Among them, Pro108Ser in 3 chymotrypsin-like protease (3CLpro) and Pro151Leu in nucleocapsid protein occurred at conserved locations among β-coronaviruses. Patients with these mutations indicated significantly lower odds ratio for developing hypoxia which required supplemental oxygen (adjusted odds ratio 0·24 [95% CI 0·07-0·88, p-value = 0·032]) after adjustments for age and sex, versus those lacking this haplotype in the canonical Clade 20B spread in Japan. The Pro108Ser 3CLpro enzyme decreases in the activity by 58%, and the hydrogen/deuterium exchange mass spectrometry reveals that mechanisms for decline-of-function involve structural perturbation at the substrate-binding region which is positioned behind and distant from the 108th amino acid residue of the enzyme.Interpretation: Viral genome sequencing in Tokyo showed that the specific mutant strain containing Pro108Ser mutation in 3CLpro, ameliorates the COVID-19 severity. This Pro108Ser mutant in 3CLpro reduces the catalytic activity of the protein by 50%. The mutant strain rapidly outcompeted pre-existing variants to become the dominant one in Japan. Our results may benefit the efforts under way to design small molecular compounds or antibodies targeting 3CLpro.Funding: Keio Gijuku Academic Development Funds and AMED (Grant Number JP20he0622043)Declaration of Interests: Authors have no conflicts of interests.Ethics Approval Statement: The study protocol was approved by the Ethics Committee of Keio University School of Medicine (approval number: 20200062).

Severity of COVID-19 is inversely correlated with increased number counts of non-synonymous mutations in Tokyo (preprint)

Background: SARS-CoV-2 genome accumulates point mutations in a constant manner. Whether the accumulation of point mutations is correlated with milder manifestations of COVID-19 remains unknown. Methods: We performed SARS-CoV-2 genome sequencing in 90 patients with COVID-19 infection treated at a tertiary medical center in Tokyo between March and August 2020. The possible association between disease severity and viral haplotype was then assessed by counting the number of mutations in addition to performing phylogenic tree analysis, comparative amino acid sequence analysis among {beta}-coronaviruses, and mathematical prediction of the functional relevance of amino acid substitutions. Results: The number of non-synonymous mutations was inversely correlated with COVID-19 severity, as defined by requiring oxygen supplementation. Phylogenic tree analysis identified two predominant groups which were separated by a set of 6 single nucleotide substitutions, including four leading to non-synonymous amino acid substitutions. Among those four, Pro108Ser in 3 chymotrypsin-like protease (3CLpro) and Pro151Leu in nucleocapsid protein occurred at conserved locations and were predicted to be deleterious. Patients with Pro108Ser in 3CLpro and Pro151Leu in nucleocapsid protein had a lower odds ratio for developing hypoxia requiring supplemental oxygen (odds ratio of 0.24 [95% confidence interval of 0.07-0.88, P-value = 0.032]) after adjustments for age and sex, compared with patients lacking this haplotype in Clade 20B. Conclusion: Viral genome sequencing in 90 patients treated in the Tokyo Metropolitan area showed that the accumulation of point mutations, including Pro108Ser in 3CLpro and Pro151Leu in nucleocapsid protein, was inversely correlated with COVID-19 severity. Further in vitro research is awaited.

Clinical Utility of SARS-CoV-2 Whole Genome Sequencing in Deciphering Source of Infection (preprint)

The novel coronavirus disease (COVID-19) pandemic caused by SARS-CoV-2 is a major threat to humans. Recently, we encountered two seemingly separate COVID-19 clusters in a tertiary care medical center. Whole viral genome sequencing detected the haplotype of the SARS-CoV-2 genome and the two clusters were successfully distinguished by the viral genome haplotype. Concurrently, there were nine COVID-19 patients clinically unlinked to clusters #1 or #2 that necessitated the determination of the source of infection. Such patients had similar haplotypes to those in cluster #2 but were devoid of two rare mutations characteristic to cluster #2. This suggested that these nine cases of "probable community infection" indeed had community infection and were not derived from cluster #2. Whole viral genome sequencing of SARS-CoV-2 is a powerful measure not only for monitoring the global trend of SARS-CoV-2 but also for identifying the source of infection of COVID-19 at a level of institution.