ABSTRACT
Background Repeated vaccination against SARS-CoV-2 increases serological response in kidney transplant recipients (KTR) with high interindividual variability. No decision support tool exists to predict SARS-CoV-2 vaccination response in KTR. Methods We developed, internally and externally validated five different multivariable prediction models of serological response after the third and fourth vaccine dose against SARS-CoV-2 in KTR. Using 27 candidate predictor variables, we applied statistical and machine learning approaches including logistic regression (LR), LASSO-regularized LR, random forest, and gradient boosted regression trees. For development and internal validation, data from 585 vaccinations were used. External validation was performed in four independent, international validation datasets comprising 191, 184, 254, and 323 vaccinations, respectively. Findings LASSO-regularized LR performed on the whole development dataset yielded a 23- and 11- variable model, respectively. External validation showed AUC-ROC of 0.855, 0.749, 0.828, and 0.787 for the sparser 11-variable model, yielding an overall performance 0.819. Interpretation An 11-variable LASSO-regularized LR model predicts vaccination response in KTR with good overall accuracy. Implemented as an online tool, it can guide decisions when choosing between different immunization strategies to improve protection against COVID-19 in KTR.
Subject(s)
COVID-19ABSTRACT
Importance Response to SARS-CoV-2 vaccines in kidney transplant recipients (KTR) is severely reduced. Heterologous 3rd vaccination combining mRNA and vector vaccines did not increase seroconversion at four weeks after vaccination but evolution of antibody levels beyond the first month remain unknown. Objective To assess changes in antibody response following a 3rd vaccination with mRNA or vector vaccine in KTR from month one to month three after vaccination. Design, Setting and Participants Three-month follow-up (pre-specified secondary endpoint) of a single-center, single-blinded, 1:1 randomized, controlled trial on 3rd vaccination against SARS-CoV-2 in 201 KTR who did not develop SARS-CoV-2 spike protein antibodies following two doses of an mRNA vaccine. Intervention(s) mRNA (BNT162b2 or mRNA-1273) or vector (Ad26COVS1) as third SARS-CoV-2 vaccine Main Outcomes and Measures Main outcome was seroconversion at the second follow-up between 60-120 days after the 3rd vaccination. Subsequently, higher cut-off levels associated with neutralizing capacity and protective immunity were applied (i.e. >15, >100, >141 and >264 BAU/mL). In addition, trajectories of antibody levels from month one to month three were analyzed. Finally, SARS-CoV-2 specific CD4 and CD8 T-cells at four weeks were compared among the 18 top responders in both groups. Results A total of 169 patients were available for the three-month follow-up. Overall, seroconversion at three months was similar between both groups (45% versus 50% for mRNA and vector group, respectively; OR=1.24, 95%CI=[0.65, 2.37], p=0.539). However, when applying higher cut-off levels, a significantly larger number of individual in the vector group reached antibody levels > 141 and > 264 BAU/mL at the three-month follow-up (141 BAU/mL: 4% vs. 15% OR=4.96, 95%CI=[1.29, 28.21], p=0.009 and 264 BAU/mL: 1% vs. 10% OR=8.75, 95%CI=[1.13, 396.17], p=0.018 for mRNA vs. vector vaccine group, respectively). In line, antibody levels in seroconverted patients further increased from month one to month three in the vector group while remaining unchanged in the mRNA group (median increase: mRNA= 1.35 U/mL and vector = 27.6 U/mL, p = 0.004). Of particular note, there was no difference in the CD4 and CD8 T-cell response between the mRNA and vector vaccine group at month one. Conclusions and Relevance: Despite a similar overall seroconversion rate at three months following third vaccination in KTR, a heterologous third booster vaccination with Ad26COVS1 resulted in significantly higher antibody levels in responders. Trial Registration: EurdraCT: 2021-002927-39
ABSTRACT
Introduction: SARS-CoV-2 gains cell entry via angiotensin-converting enzyme (ACE) 2, a membrane-bound enzyme of the “alternative” (alt) renin-angiotensin system (RAS). ACE2 counteracts angiotensin II by converting it to potentially protective angiotensin 1–7.Methods Using mass spectrometry, we assessed key metabolites of the classical RAS (angiotensins I–II) and alt-RAS (angiotensins 1–7 and 1–5) pathways as well as ACE and ACE2 concentrations in 159 patients hospitalized with COVID-19, stratified by disease severity (severe, n = 76; non-severe: n = 83). Plasma renin activity (PRA-S) was calculated as the sum of RAS metabolites. We estimated ACE activity using the angiotensin II:I ratio (ACE-S) and estimated systemic alt-RAS activation using the ratio of alt-RAS axis metabolites to PRA-S (ALT-S). We applied mixed linear models to assess how PRA-S and ACE/ACE2 concentrations affected ALT-S, ACE-S, and angiotensins II and 1–7.Results Median angiotensin I and II levels were higher with severe versus non-severe COVID-19 (both p < 0.05), demonstrating activation of classical RAS. The difference disappeared with analysis limited to patients not taking a RAS inhibitor. ALT-S in severe COVID-19 increased with time (days 1–6: 0.12; days 11–16: 0.22) and correlated with ACE2 concentration (r = 0.831). ACE-S was lower in severe versus non-severe COVID-19 (p < 0.001), but ACE concentrations were similar between groups and weakly correlated with ACE-S (r = 0.232). ACE2 and ACE-S trajectories in severe COVID-19, however, did not differ between survivors and non-survivors. Overall RAS alteration in severe COVID-19 resembled severity of disease-matched patients with influenza. In mixed linear models, renin activity most strongly predicted angiotensin II and 1–7 levels. ACE2 also predicted angiotensin 1–7 levels and ALT-S. No single factor or the combined model, however, could fully explain ACE-S. ACE2 and ACE-S trajectories in severe COVID-19 did not differ between survivors and non-survivors.Conclusions Angiotensin II was elevated in severe COVID-19 but markedly influenced by RAS inhibitors and driven by overall RAS activation. ACE-S was significantly lower with severe COVID-19 and did not correlate with ACE concentrations. A shift to the alt-RAS axis because of increased ACE2 could partially explain the relative reduction in angiotensin II levels.