Neutralising antibody responses to emerging SARS‐CoV‐2 omicron sub‐variants are enhanced in persons with hybrid immunity Authors: Sharon L Walmsley, Majid Nabipoor, Freda Qi, Leif Erik Lovblom, Rizani Ravindran, Karen Colwill, Anne Claude Gingras Published: 2026 ABSTRACT AbstractObjectivesTo assess neutralizing antibody responses to emerging SARS‐CoV‐2 variantsMethodsWe explored SARS‐CoV‐2 neutralizing antibodies in140 adults participating in a longitudinal study of antibody responses to COVID vaccines. Participants reported monthly on vaccine doses, and SARS‐CoV‐2 infections. Serum was collected in December 2023 and April 2024 to determine neutralizing antibody to Wuhan ancestral strain and Omicron XBB.1.5 and JN.1 subvariants. Multivariate analysis assessed protective factors against a new SARS‐CoV‐2 infection between the two sampling times and 10 months follow up.Results74% had hybrid immunity and 26% vaccine only immunity at study entry. A neutralizing antibody response was detected in 99% to Wuhan ancestral strain compared to 86% and 79% respectively to Omicron XBB.1.5 and JN.1 subvariants. For those with detectable antibodies the levels decayed, faster in those with vaccine only immunity and to the JN.1 subvariant. Neither the presence or the level of neutralizing antibody to JN.1 was associated with protection with a new SARS‐Cov‐2 infection over the 10‐month follow up period.ConclusionNeutralizing antibody levels to Omicron JN.1 subvariant were identified in 79% overall but in only 56% with vaccine only immunity. The presence or level was not associated with protection form a new SARS‐CoV‐2 infection over 10‐months. This data supports the development of pan‐coronavirus vaccine to expand protection against emerging variants. Introduction Despite the efficacy of vaccines in controlling the COVID‐19 pandemic, 1 mutations in the SARS‐CoV‐2 genome have led to the emergence of new variants with greater transmissibility and enhanced immune evasion. 2 , 3 Vaccines targeted to the new variants and sub‐variants have been developed but may not provide sufficient cross‐reactive protection. 4 Further, the virus has often mutated further before significant population uptake of the new vaccine boosters occurs. 5 Improved understanding of the vaccine and viral induced immune responses is important for optimising vaccine strategies. 6 Numerous studies have shown that hybrid immunity (SARS‐CoV‐2 infection plus vaccine‐induced immunity) provides greater protection against new SARS‐CoV‐2 variant infections compared to vaccine‐only immunity but the correlates of the improved response remain unclear 7 , 8 , 9 , 10 . Although hybrid immunity results in increased binding and neutralising antibody levels compared to vaccine‐only immunity, a threshold value for protection from a new infection 8 , 9 , 11 , 12 , 13 , 14 , 15 has not been determined. In previous work, we were unable to demonstrate that the presence of neutralising antibodies to the Omicron sub‐variant XBB.1.5 was associated with protection from infection during the subsequent Omicron waves. 16 We extend our work to address the generation of neutralising antibodies to the SARS‐CoV‐2 Omicron subvariant JN.1 and study whether it protects against new SARS‐CoV‐2 infections during the KP.3 sub variant wave of infection in Ontario. Results Of the 180 contacted 140 consented to this sub‐study, 104 (74.3%) had hybrid immunity (80 from the ≥ 70 years age cohort and 24 from the 30–50 years cohort) while 36 (25.7%) had vaccine‐only immunity (27 from the older age cohort and 9 from the younger cohort). As reported previously, 17 the hybrid and vaccine‐only immunity groups were similar in gender (approximately 75% women), BMI (over 25% with obesity), racial background (over 90% Caucasian), smoking status and comorbidities (over 50% with comorbidity). The participants in this study were representative of those in the parent STOPCoV study. 18 Their characteristics are shown in Supplementary table 1. The average number of vaccines received up to the baseline of this sub‐study differed between the hybrid immunity (five doses) and vaccine‐only immunity (six doses) groups (P‐value = 0.009). The time from the last SARS‐CoV‐2 infection to baseline was a mean of 358 days in the hybrid group, and the time from the last vaccine dose to baseline was a mean of 148 days in the vaccine‐only group (P < 0.0001 for comparison of times). During the 5‐month follow‐up, 12 (12%) of the hybrid immunity group and eight (22%) of the vaccine‐only immunity group received an additional vaccine dose (mRNA encoding Kp.2). Additionally, during the 5‐month follow‐up, 10 (9.6%) participants in the hybrid immunity group and five (13.9%) participants in the vaccine‐only immunity group experienced a new SARS‐CoV‐2 infection (during the time when the Kp.3 sub‐variant was dominating). Baseline analysis A neutralising antibody response to the Wuhan ancestral strain at baseline was detected in 138 participants (98.6%). For those with a detectable neutralising antibody, those in the hybrid cohort developed a significantly higher mean spike neutralising antibody level of 4.35 (log10 (1/ID50)) compared to 3.87 (log10 (1/ID50)) in the vaccine‐only cohort (P < 0.0001), as shown in Table 1. Table 1The proportion of participants with detectable neutralising antibodies and levels of these antibodies for Wuhan ancestral strain and Omicron XBB.1.5 and JN.1 sub‐variants, stratified by hybrid or vaccine‐only immunity at baselineCharacteristicsHybrid immunityVaccine‐only immunityTotal P‐valueCohort, n 10436140 – Wuhan: Neutralising antibody, n (%)Non‐detectable‐2 (5.6%)2 (1.4%)–Detectable104 (100%)34 (94.4%)138 (98.6%)Neutralising antibody levels log10 (1/ID50)Mean (SD)4.35 (0.46)3.87 (0.68)4.23 (0.56)< 0.0001Median (IQR)4.34 (4.12, 4.63)3.73 (3.50, 4.41)4.28 (3.95, 4.56)< 0.0001Omicron XBB.1.5: Neutralising antibody, n (%)Non‐detectable6 (5.8%)14 (38.9%)20 (14.3%)< 0.0001Detectable98 (94.2%)22 (61.1%)120 (85.7%)Neutralising antibody levels log10 (1/ID50)Mean (SD)3.17 (0.59)2.68 (0.60)3.08 (0.62)0.0006Median (IQR)3.34 (2.84, 3.57)2.51 (2.24, 3.09)3.19 (2.54, 3.52)0.0012Omicron JN.1: Neutralising antibody, n (%)Non‐detectable13 (12.5%)16 (44.4%)29 (20.7%)0.0001Detectable91 (87.5%)20 (55.6%)111 (79.3%)Neutralising antibody levels Log10 (1/ID50)Mean (SD)2.99 (0.61)2.44 (0.55)2.89 (0.64)0.0003Median (IQR)3.09 (2.51, 3.47)2.27 (2.01, 2.70)2.84 (2.41, 3.39)0.0005ID50, half maximal inhibitory dilution; IQR, interquartile range; n, number; SD, standard deviation. The proportion of participants with detectable neutralising antibodies and levels of these antibodies for Wuhan ancestral strain and Omicron XBB.1.5 and JN.1 sub‐variants, stratified by hybrid or vaccine‐only immunity at baseline ID50, half maximal inhibitory dilution; IQR, interquartile range; n, number; SD, standard deviation. The proportion of participants with detectable neutralising antibodies was lower for the XBB.1.5 and JN.1 sub‐variants than for the Wuhan ancestral strain (86% and 79%, respectively). Participants with hybrid immunity had detectable neutralising antibodies to the Omicron XBB.1.5 (94%) and JN.1 (88%) sub‐variants, compared to 61% and 56%, respectively, of those with vaccine‐only immunity. Similarly, the antibody levels for those with detectable neutralising antibody to both sub‐variants were higher in those with hybrid immunity. When considering potential covariates, hybrid immunity was the most significant predictor of neutralising antibody detection for Omicron XBB.1.5 and JN.1 in logistic regression analysis. The odds of detecting antibodies in individuals with vaccine‐only immunity were significantly lower against Omicron XBB.1.5 (adjusted OR [aOR]: 0.07; 95% CI: 0.02–0.22) and JN.1 (aOR: 0.13; 95% CI: 0.05–0.33). Using logistic and linear regression, age, gender, comorbidity and the time from last infection or vaccine dose were not associated with either the detection or the level of neutralising antibody to either of the sub‐variants (Supplementary table 2). The number of vaccinations up to baseline significantly increased neutralising antibody levels for XBB.1.5 (adjusted β [aβ]: 0.18; 95% CI: 0.07, 0.29) and for JN.1 (aβ: 0.18; 95% CI: 0.06, 0.29), but had no significant effect on antibody levels against Wuhan ancestral strain. Baseline to 5‐month analysis Overall, 134 participants provided specimens at the follow‐up visit. Figure 1 illustrates the neutralising antibody levels at baseline of this sub‐study and the 5‐month follow‐up for the Wuhan ancestral strain, Omicron XBB.1.5 and JN.1 sub‐variant strains. For participants with detectable neutralising antibody, those with hybrid immunity demonstrated higher antibody levels across all three strains at baseline. The proportion of participants who did not have detectable neutralising antibodies at the follow‐up increased with the new sub‐variants, rising from 0.7% for Wuhan ancestral strain to 13% for Omicron XBB.1.5 and 23% for JN.1. For the JN.1 sub‐variant, 86% with hybrid immunity but only 52% with vaccine‐only immunity had detectable neutralising antibodies. For those with detectable neutralising antibodies, the levels were higher for those with hybrid immunity to Wuhan ancestral strain and Omicron XBB.1.5 relative to those with vaccine only immunity. The neutralising antibody levels to JN.1 were lower overall and did not differ between those with hybrid immunity relative to vaccine‐only (Table 2). The neutralising antibody levels log 10 (1/ID 50) at the baseline and 5‐month follow‐up for the Wuhan ancestral strain (left), Omicron XBB.1.5 (middle) and Omicron JN.1 (right) variant strains, stratified by hybrid/vaccine‐only immunity. Table 2The proportion of participants with detectable neutralising antibodies and levels of these antibodies for Wuhan ancestral strain, and Omicron XBB.1.5 and JN.1 sub‐variants, at the 5 month follow up stratified by hybrid/vaccine‐only immunity at baselineHybrid immunityVaccine‐only immunityTotal P‐valueCohort, n 9935134–Wuhan: Neutralising antibody, n (%)Non‐detectable0 (0.0%)1 (2.9%)1 (0.7%)0.0914Detectable99 (100.0%)34 (97.1%)133 (99.3%)Neutralising antibody levels log10 (1/ID50)Mean (SD)4.21 (0.44)3.73 (0.58)4.09 (0.52)< 0.0001Median (IQR)4.24 (3.88, 4.58)3.66 (3.38, 4.03)4.11 (3.71, 4.48)< 0.0001Omicron XBB.1.5: Neutralising antibody, n (%)Non‐detectable7 (7.1%)11 (31.4%)18 (13.4%)0.0003Detectable92 (92.9%)24 (68.6%)116 (86.6%)Neutralising antibody levels log10 (1/ID50)Mean (SD)3.11 (0.55)2.85 (0.54)3.06 (0.56)0.0353Median (IQR)3.15 (2.77, 3.51)2.87 (2.36, 3.26)3.06 (2.71, 3.44)0.0419Omicron JN.1: Neutralising antibody, n (%)Non‐detectable14 (14.1%)17 (48.6%)31 (23.1%)< 0.0001Detectable85 (85.9%)18 (51.4%)103 (76.9%)Neutralising antibody levels log10 (1/ID 50)Mean (SD)2.83 (0.55)2.64 (0.50)2.80 (0.54)0.1722Median (IQR)2.80 (2.48, 3.23)2.68 (2.19, 3.08)2.75 (2.40, 3.22)0.1957ID50, Half maximal inhibitory dilution; IQR, interquartile range; n, number; SD, standard deviation. The proportion of participants with detectable neutralising antibodies and levels of these antibodies for Wuhan ancestral strain, and Omicron XBB.1.5 and JN.1 sub‐variants, at the 5 month follow up stratified by hybrid/vaccine‐only immunity at baseline ID50, Half maximal inhibitory dilution; IQR, interquartile range; n, number; SD, standard deviation. In ordinal logistic regression, vaccination during the 5‐month follow‐up period was associated with higher odds of achieving higher neutralising antibody levels across all variants: Wuhan ancestral strain (aOR: 6.24; 95% CI: 2.11–18.50), Omicron XBB.1.5 (aOR: 8.21; 95% CI: 3.42–21.11) and Omicron JN.1 (aOR: 6.59; 95% CI: 2.77–15.96). New SARS‐CoV‐2 infections during this period were associated with higher odds of achieving elevated neutralising antibody levels for the Wuhan ancestral strain (aOR: 4.03; 95% CI: 1.23–13.24), Omicron XBB.1.5 (aOR: 4.34; 95% CI: 1.75–11.20) and Omicron JN.1 (aOR: 7.51; 95% CI: 2.88–19.77). The data are displayed in Supplementary table 3. The decay slopes of neutralising antibody levels for the 102 participants who did not receive an additional vaccine dose or a new SARS‐CoV‐2 infection between the sampling times was determined. Neutralising antibody levels decayed faster for the vaccine‐only group relative to the hybrid immunity group and were faster for JN.1 overall. The decay slope for neutralising antibodies to Wuhan ancestral strain, Omicron XBB.1.5 and Omicron JN.1 were −0.2, −0.21 and −0.37, respectively, for those with hybrid immunity and −0.24, 0.35, and −0.68 for those with vaccine‐only immunity. Protection against new SARS CoV‐2 infection During the 5‐month follow‐up period, 15 new SARS‐CoV‐2 infections were recorded, increasing to 25 by the 10‐month follow‐up period ending on 8 October 2024. Among the 104 participants with hybrid immunity at the baseline visit, 10 (9.6%) experienced a new SARS‐COV‐2 infection during the 5‐month follow‐up, and 16 (15.4%) during the 10‐month follow‐up. Of the 36 participants with vaccine‐only immunity at baseline, five (13.9%) experienced a new SARS‐CoV‐2 infection during the 5‐month follow‐up, and 92 (25%) during the 10‐month follow‐up. Twenty participants received additional vaccine doses during the 5‐month follow‐up, and 52 participants received additional vaccine doses during the 10‐month follow‐up (Supplementary table 1). In the multivariable analysis, none of age cohort, gender, obesity, presence of comorbidity nor receipt of an additional vaccine dose were associated with protection from a new SARS‐CoV‐2 infection in either the 5 or 10‐month follow up period (Figure 2). Further, the level of JN.1 neutralising antibody at the baseline visit was not associated with protection from a new SARS‐CoV‐2 infection at either the 5‐ or 10‐month follow up period. In a separate model, the presence of JN.1 antibodies was also not associated with protection from a new SARS‐CoV‐2 infection. Univariate and Multivariate analysis for protection against a new SARS‐CoV‐2 infection over the short‐term (5 month) and long‐term (10 month) follow‐up period. BMI, body mass index; FU, follow up; M, month. Those with undetectable antibodies to JN.1 were assigned zero as the antibody level. Discussion Neutralising antibodies to the SARS‐CoV‐2 JN.1 sub‐variant were identified in 79% of participants in a cohort of 140 heavily vaccinated persons and were more likely detected in those with hybrid immunity (88%) compared to those with vaccine‐only immunity (56%). These percentages were lower than the proportion with detectable neutralising antibodies to the Wuhan ancestral strain (100% and 94%, respectively) and to the Omicron XBB.1.5 sub‐variant (94% and 61%). This neutralising antibody study was performed in December 2023–April 2024 with most participants receiving the initial two dose mRNA vaccine in summer of 2021 and the XBB.1.5 booster in the fall of 2023. In those with detectable antibodies, the neutralising antibody levels were higher in those who had received more COVID‐19 vaccines prior to testing. Of those with detectable antibodies, the levels in both groups were lower against the JN.1 sub‐variant than either Wuhan ancestral strain or Omicron XBB.1.5. The levels of the antibodies declined in all groups over a 5‐month period. The detection of neutralising antibodies to the JN.1 sub‐variant was not associated with protection against a new SARS‐CoV‐2 infection during the subsequent 10 months of observation when the Omicron Kp.3 sub‐variant was dominant. The JN.1 sub‐variant of COVID‐19 emerged as the dominant strain worldwide at the end of 2023. 19 This variant harbours a unique combination of mutations inherited from the BA.2.86 lineage, which have been linked to increased transmissibility and enhanced immune evasion in part because of increased binding affinity to the angiotensin‐converting enzyme 2 receptor. 19 Reduced cross neutralisation of the XBB.1.5 vaccine have been observed. 20 , 21 , 22 The dominant JN.1 subvariant was gradually replaced with Omicron subvariants particularly KP.2 and KP.3. 23 Our findings align with previous research indicating the limited breadth of neutralising antibody responses after mRNA vaccination. 22 The failure to detect neutralising antibodies to the JN.1 variant in 44% of those with vaccine‐only immunity raises further concern about the emergence of viral variants capable of overcoming population immunity derived from prior infection or vaccine. 24 Our population was very diligent with obtaining the new vaccine boosters including the XBB.1.5 booster as they were recommended by public health authorities. This observation adds further support to the concern and challenges in manufacturing keeping up with viral diversity and calls for a pan‐corona vaccine. 25 We also demonstrated that the neutralising antibody responses decayed over a 5‐month period again pointing to the potential limited duration of protection that may be afforded. Although there has been concern that the elderly may not respond as well to vaccines 26 and be at higher risk of infection with the new variants we did not find that age was predictive of neutralising antibody responses to JN.1. In our regression models, age was not correlated with either neutralising antibody detection nor with protection from a new SARS‐CoV‐2 infection. In this study, we did not find that neutralising antibodies to the Omicron JN.1 sub‐variant were protective against infection over the next 10 months when further sub‐variants, notably KP.3, were circulating. Of note, all our participants had either received the XBB.1.5 vaccine or had a SARS‐CoV‐2 infection with that sub‐variant. This further points to the continued need to update COVID vaccines to match the circulating variants. Arora and others recently demonstrated that the JN.1 adapted mRNA vaccine bretovameran (BioNTech‐Pfizer, Mainz, Germany) boosted neutralising activity to Kp.3 although less well than to JN.1. 23 A key strength of our study is the longitudinal follow‐up with prospective data collection on vaccine doses and SARS‐CoV‐2 infection, allowing us to compare neutralising antibody responses in individuals with and without prior infection. Vaccine uptake including the XBB1.5 booster was high. Our cohort is unique in that they have participated in a decentralised longitudinal study in which they are well characterised and includes a high proportion of elderly persons. We demonstrated a decline in neutralising antibody levels to JN.1 with time, but as we only had one follow up measure at 5 months were unable to capture longer term data particularly whether antibody levels were lower at the time of a new SARS‐CoV‐2 infection. Conclusion Neutralising antibodies to Omicron subvariant JN.1 were identified in 79% of our cohort, but in only 56% of those with vaccine‐only immunity. The proportion with detectable antibodies and their levels were lower than for Wuhan ancestral strain or the Omicron XBB.1.5 sub‐variant. The presence or level of neutralising antibodies was not associated with protection from a new SARS‐CoV‐2 infection during the subsequent 10‐month follow‐up. These observations further point to the need for continuing genomic surveillance, regular vaccine updates but more importantly the need for vaccines targeting conserved epitopes to enable pan‐coronavirus protection. Methods Parent study The parent prospective decentralised study (www.STOPCoV.ca) enrolled 1286 participants in 2021. 18 Its main objective was to compare the antibody responses to COVID‐19 vaccines (IgG antibody levels (BAU/mL‐binding antibody units per millilitre)) to the spike trimer, its receptor binding domain (RBD) and nucleocapsid protein 27 in an older relative to a younger cohort performed by enzyme‐linked immune absorbant assay (ELISA) on 3‐monthly self‐collected dried blood spots (DBS). New SARS‐CoV‐2 infections were documented to October 2024 based on monthly e‐reports of a positive PCR or rapid antigen test (RAT), or the development of nucleocapsid antibodies above the seropositivity threshold 12 from the DBS collected as part of the parent study. Sub‐study participant recruitment and consent In the fall of 2022, email invites were sent to a random selection of 180 STOPCoV participants who had consented to be contacted for future to participate in a sub‐study of neutralising antibodies to Omicron XBB.1.5 sub‐variant. 16 The same participants were re‐contacted in the fall of 2023 to e‐consent to this current sub‐study to assess neutralising antibodies against Wuhan ancestral strain and the Omicron XBB.1.5 and JN.1 sub‐variants. Participants classified as having vaccine‐only immunity in the fall of 2022 were reclassified as having hybrid immunity in the fall of 2023 as necessary. This sub‐study was approved by the University Health Network (UHN) research ethics board (CAPCR Study 22–5715) and was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Sample collection With consent, blood collection kits with barcoded containers were sent to the enrolled participants. An external lab collected, packaged and shipped specimens to study team at the UHN at baseline (December 2023) and at an average of 5 months later (April 2024). Separated serum was transferred to the Lunenfeld‐Tanenbaum Research Institute (LTRI) at Sinai Health for further processing (Sinai REB study 23‐0069‐E). Laboratory analysis and reporting The lentivirus neutralisation assays were performed to assess neutralisation capacity against Wuhan ancestral strain SARS‐CoV‐2 and Omicron sub‐variants including XBB.1.5 and JN.1. The generation of spike‐pseudotyped lentivirus particles was performed as described previously with minor modifications. 16 , 27 , 28 Relative luciferase units were first normalised to positive and negative controls. A sigmoidal curve with the association Y = 100/(1 + 10((Log10ID50 − x) × HillSlope)) was fitted and an ID50 value was calculated for each sample, using the nlsLM function from the minpack.lm package in R. 29 , 30 Samples that did not show neutralisation capacity at the serum dilutions performed were assigned an ID50 value of 1. Statistical analysis Cohort characteristics based on hybrid or vaccine‐only immunity were compared at the baseline visit. Neutralising antibody log10 (1/ID50), for the Wuhan ancestral strain, Omicron XBB.1.5, and JN.1 sub‐variants were analysed at baseline and at the follow‐up sampling time. First, we determined the proportion of those with neutralising antibody levels above the threshold level, which we designated as detectable or not. Bayesian logistic regression examined the association between the detection of neutralising antibody to each variant and the covariates age cohort, gender, body mass index (BMI), presence of comorbidity (any one of cardiovascular, respiratory disease, diabetes, cancer), hybrid vs vaccine only immunity, the number of vaccine doses received prior to the baseline visit and time since last SARS CoV‐2 infection or vaccination. Second, for those with detectable antibodies, we compared the levels (mean and 95% confidence intervals) and used linear regression with a similar covariate structure to the logistic regression model. The neutralising antibody levels from baseline to the 5‐month follow‐up were compared to evaluate the trajectory of the response. Those with undetectable antibody were given the value of zero. Bayesian ordinal logistic regression identified predictors of change. A descriptive analysis assessed the decay slopes (i.e. the difference in log10 (1/ID50) of the levels between the two visits) of neutralising antibody levels among participants with vaccine only versus hybrid immunity who neither experienced a new SARS‐CoV‐2 infection nor received an additional vaccine dose during the 5‐month follow‐up. Cox regression analysis was used to identify factors associated with protection from a new SARS‐CoV‐2 infection. The outcome was the time to the first SARS‐CoV‐2 infection, measured from the baseline visit date to the 5‐month sampling time (short‐term) and from the baseline visit date to the 10‐month follow‐up (long‐term), which was until 8 October 2024. The analysis for both periods incorporated fixed‐time independent predictors, including age cohort, gender, BMI, comorbidity, hybrid or vaccine‐only immunity at baseline, JN.1 antibody levels at the baseline visit and receipt of an additional vaccine dose between the time periods. JN.1 levels were treated as dichotomous (detectable antibody, yes or no) or continuous (levels, inputting the value of zero for those with undetectable antibody) in separate models. A significance threshold of α = 0.05 was applied for all statistical hypothesis tests. Data manipulation and analyses were conducted using SAS (version 9.4) and R (version 4.2.2). Author contributions Conceptualisation: SLW, ACG, and KC. Methodology: SLW and KC. Validation: SLW, LEL, MN, FQ, and KC. Formal analysis: MN and LEL. Investigation: RR and FQ. Resources: SLW and ACG. Data curation: SLW, MN, LEL, KC and ACG. Writing – original draft preparation: SLW and MN. Writing – review and editing: SLW, MN, LEL, KC and ACG. Visualisation: SLW, MN. Supervision: SLW and ACG. Project administration: RR. Funding acquisition: SLW and ACG. All authors have read and agreed to the published version of the manuscript. Conflict of interest SLW served on advisory boards, spoken at CME events and has received funding for investigator driven research projects from ViiV Healthcare, Merck and Gilead but has no conflicts directly related to this work. KC has reported honoraria for presentation to the Ontario Society of Clinical Chemists and has no conflicts directly related to this work. ACG reports funding from Providence therapeutics and it is not related to this work. All other authors have no conflicts of interest to disclose. Supporting information Supplementary table 1 Supplementary table 2 Supplementary table 3