Association of Lower Antispike Antibody Levels with Mortality in ICU Patients with COVID-19 Disease

Background Though vaccines have been reported as highly efficacious in preventing severe COVID-19 disease, there is emerging data of severe infections, albeit a small number, in vaccinated individuals. We have conducted a retrospective observational study to assess the clinical characteristics, immunological response, and disease outcomes among the vaccinated and unvaccinated patients admitted to the ICU with severe COVID-19 disease. Methods Study Design and Participants. We conducted a retrospective observational study in COVID ICU of a tertiary care hospital. Data were collected from the month of 1 April 2021 to 31 November 2021. All adult patients admitted to the ICU having severe COVID-19 disease were included in the study. Data were collected from the medical records database which included demographics, a clinical course in the ICU, laboratory and radiological parameters, and disease outcomes. In a subset of patients, cell-mediated immunity and S1S2-neutralising antibody assessment was done. Results A total of 419 patients with severe COVID-19 were included in the study. Of the 419 patients, 90 (21.5%) were vaccinated, and 329 (78.5%) were unvaccinated. There was a significantly higher mortality in unvaccinated severe COVID 19 patients as compared to vaccinated severe COVID patients (46.2% vs 34.4%; P < 0.0455). The neutralizing antibody titre was significantly higher in survivors as compared to nonsurvivors (2139.8, SE ± 713.3 vs 471, SE ± 154.4); P < 0.026. Conclusion Our study suggests the association of lower neutralizing antibody levels with mortality in ICU patients admitted with COVID-19 breakthrough infections.


Introduction
Vaccines have been designed and rolled out to protect against severe COVID-19 disease and to reduce mortality. India initiated its vaccination program in the month of January, 2021. Oxford Astrazeneca Covidshield vaccine manufactured by Serum Institute of India and Covaxin (BBV152) developed by Bharat Biotech in collaboration with the Indian Council of Medical Research were the two vaccines initially available for mass vaccination. Covaxin is a whole inactivated virus vaccine. In a phase 3 trial, BBV152 was found to have overall efcacy of 77.8% (95% CI, 65.2%-86.4%) in preventing symptomatic disease [1]. As of 31 January 2020, Covaxin has been granted emergency use approval in 13 countries. It is recommended for individuals more than 6 years of age given as two doses 28 days apart. Covishield is a recombinant replication-defcient chimpanzee adenovirus vector vaccine containing genetic material of spike protein. It has an efcacy of 72% against symptomatic SARS-CoV-2 infection and is administered as 2 doses, 12 to 16 weeks apart in individuals more than 12 years of age [2]. For both Covishield and Covaxin, booster dose is recommended after 4 to 6 months.
Another vaccine approved in India is Sputnik V produced by Gamaleya Institute in Russia. It is a viral vectorbased vaccine taken in 2 doses 21 days apart. Sputnik V was reported to have higher efcacy in phase 3 trial approaching that of mRNA vaccines [3]. Pfzer/biotech and Moderna were the foremost vaccines introduced with a reported vaccine efcacy of 95% against symptomatic disease [4]. Tey are mRNA vaccines, but usage in India seemed difcult because of the storage temperature required (must be stored at freezer-level temperature).
In spite of successful vaccination programs covering the population in general and at-risk individuals in particular, breakthrough infections causing severe disease and ICU admission have been reported [5][6][7][8]. Efcacy of the vaccines needs to be continuously monitored for efcient implementation of the vaccination program to minimize the severity of the disease and mortality [9]. Clinical data emanating from the real world based on the vaccination status of patients with severe disease admitted to ICU is sparse and less reported in the literature. We therefore conducted a retrospective, observational study to assess the clinical outcomes of ICU patients based on the vaccination status.

Patients and Methods
Tis is a single-centre retrospective, observational study conducted in the medical ICU of a tertiary care hospital in south India. Data were collected from RT-PCR-positive patients (>18 years of age) admitted to ICU with severe COVID-19 disease (n � 419) between April 2021 to November 2021, which included both vaccinated and unvaccinated patients. Exclusion criteria were those with unknown vaccination status, pregnant women, and patients or relatives who declined to consent to participate in the study.
All the patients received treatment as per standard protocols. Data were collected from the medical records database which included demographic characteristics, comorbidities, type of vaccine, infammatory (D-dimer, ferritin, CRP, IL6, and LDH), hematological and biochemical parameters clinical course in the ICU, and disease outcome. Te study was approved by the Institutional Ethics Committee of AIG Hospitals, Hyderabad, India.
In a subset of individuals, the immunological assessment was performed by evaluating neutralizing (antispike or S1S2) antibodies and enumeration of T and B cell responses. S1S2 IgG antibodies were enumerated on automated Diasorin Liaison XL by Chemiluminescence Immunoassay (CLIA) [10,11]. Te detection limit is ≥3.8 AU/ML, and the samples with ≥15 AU/ml were considered positive for neutralizing antibodies. For evaluating immune responses, peripheral mononuclear cells (PBMCs) were isolated using Hisep (Himedia) employing standard protocol. PBMCs were characterized phenotypically by fow cytometry for T and B lymphocytes [12]. CD3 (APC-H7), CD4 (Percp Cy 5.5), and CD8 (FITC) for T lymphocyte and CD20 (PE) for B lymphocyte were characterized [13]. Appropriate isotypematched, nonreactive fuorochrome-conjugated antibodies were employed as controls. Analysis of cell populations was performed by means of direct immunofuorescence with a FACS Lyrics fow cytometry (BD), and data were calculated.

Defnitions.
A breakthrough infection was defned as the detection of SARS-CoV-2 on RT-PCR assay performed at hospital admission in vaccinated patients. According to CDC, people are considered fully vaccinated for COVID-19 ≥ 2 weeks after they have received the second dose in a 2-dose series (ChadOxnCoV or Covaxin), and unvaccinated people refer to individuals that have not completed a vaccination series or received a single-dose of concerned vaccine. Te disease severity was defned as per the current Indian Council Medical Research (ICMR) guidelines (Clinical Guidance for Management of Adult COVID-19 Patients (icmr.gov.in)) [14]. Severe disease at admission was defned as patients with SpO2 < 90% or requirement of ICU care. Mortality parameters include in-hospital as well as patients discharged against medical advice with poor outcome.

Statistical Analysis.
Te baseline characteristic of the two groups was presented as mean and standard error (SE) for continuous variables. Te categorical variables were presented as % (percentage) of frequency distribution. Student's t-test, chi-square test, and Fisher's exact test were used for comparison between two groups. Univariate and multivariate logistic regression analysis was applied to see the efect of risk factors on survival status. Te risk factors in multiple logistic regression analysis were entered by enter method. Te receiver operating characteristics curve (ROC) analysis was applied to see the efect of immunological factors on survival status in subgroup analysis. Te statistical package for social sciences (SPSS 21st version) and Med Cal C were used for the analysis. A P value < 0.05 was considered statistically signifcant.

Outcomes.
Te primary outcome of the study was to compare 30-day mortality in vaccinated with unvaccinated severe COVID-19 disease patients in the ICU setting. Secondary outcomes included requirement of mechanical ventilation, vasopressor requirement, need for renal replacement therapy, and immunological assessment.

Demographics.
A total of 419 patients were admitted to the ICU with severe COVID-19 disease during the study period. Of these, 90 (21.5%) patients were vaccinated, and 329 (78.5%) patients were unvaccinated. Te baseline characteristics of the patients are depicted in Table 1. A signifcant diference (P � 0.0001) in the mean age was noted between the vaccinated and unvaccinated groups (61.9 ± 12.18 Vs. 54.4 ± 14.09 years) ( Table 1). Both the groups were predominantly comprised of males with no signifcant diference between the vaccinated and unvaccinated groups (66.7% Vs. 70.8%; P � 0.45). Majority of the patients had diabetes and hypertension in both the groups (Table 1). However, there was no signifcant difference in the number of individuals with diabetes mellitus in the vaccinated and unvaccinated groups (60% Vs. 57.8%; P < 0.703). Likewise, there was no signifcant diference in hypertension between the vaccinated and the unvaccinated groups (65.6% vs 57.8%; P < 0.180). Whole genomic sequencing of the SARS-CoV-2 virus identifed that the strains were of the delta variant.

Clinical Outcomes in ICU Patients.
Tere was signifcantly (P < 0.04) higher mortality in unvaccinated patients (152/329; 46.2%) as compared to vaccinated (31/90; 34.4%) patients with severe COVID-19. Te vaccinated individuals had received either ChadOxnCoV (58.17%) or Covaxin (38.89%). Tere was no signifcant diference in mortality (P < 0.374) between the ChadOxnCoV (25.64%) or Covaxin (35.71%) groups. Tere was no signifcant diference in invasive ventilation requirement, vasopressor requirement, and the renal replacement therapy requirement between the vaccinated and unvaccinated groups as shown in Figure 1(b). Te hospital length of stay was higher in vaccinated individuals as compared to unvaccinated individuals (17.3 vs 13.6, P < 0.001). As shown in Figure 1(a), a signifcantly higher proportion of individuals in the unvaccinated group had fever, shortness of breath (SOB), fatigue, and rhinitis as compared to the vaccinated patients with severe COVID-19 admitted to ICU.

Infammatory Markers.
Infammatory markers such as D-dimer, ferritin, lactate dehydrogenase (LDH), and C reactive protein (CRP) were measured in these patients on admission to ICU. Tere was no signifcant diference in the D-dimer (1055 vs 1190.4, P < 0.485) and ferritin (900.3 vs 924.9, P < 0.847) levels among the vaccinated and unvaccinated groups ( Table 2). LDH (902.1 vs 1024.7, P < 0.044) was found to be higher in the unvaccinated as compared to vaccinated groups. But CRP (116.9 vs 82, P < 0.00) was found to be higher in vaccinated group as compared to unvaccinated group.
However, there were 5 vaccinated patients who died despite having high neutralizing antibody levels, ranging from 658 AU/ML to 4225 AU/ML (details in Table 4).

Discussion
Global initiation of the vaccination programs has reduced the incidence of symptomatic disease, hospitalization, and death due to COVID-19 disease. Data pertaining to the clinical characteristics and outcomes in individuals developing severe disease as a result of breakthrough infection is limited. Our study describes the clinical and immunological parameters in patients admitted to ICU with severe COVID-19 after vaccination with either ChadOxnCoV-19 or Covaxin.
Higher mean age of the vaccinated group as compared to the unvaccinated group is likely due to the prioritization of elderly in the initial phases of the vaccination schedule by the Government of India. However, it is interesting to note that the number of vaccinated individuals hospitalized with severe disease is relatively lesser, and mortality is also lesser compared to the unvaccinated group in spite of the higher age during the study period pointing to the critical protective role of the vaccines against severe disease. Te incidence of comorbidities was similar in the vaccinated and unvaccinated groups, but mortality was lower in vaccinated group. We found no signifcant diference in the need for mechanical ventilation, vasopressor, and RRT between the vaccinated and unvaccinated groups which may be attributable to comorbidities contributing to severe disease and organ dysfunction. It is intriguing to note that hospital length of stay was higher in vaccinated which is probably due to the reason that individuals in vaccinated group were elderly. However, these results have to be interpreted in light of data emanating from a single center. In our cohort of unvaccinated patients admitted to ICU, the mortality was 46.2%. Tis is in agreement with the reported literature by Xie and Raquel et al. where mortality rates were between 38% and 64% in admitted ICU patients who were unvaccinated [15,16]. However, interestingly, vaccination resulted in a signifcant decrease in mortality (34.4%, P < 0.045). We found no signifcant diference in mortality among the vaccinated ICU patients with regard to the two predominant vaccines being used in India. Tis   indicates all vaccines protect against mortality even after individuals have developed severe diseases [17]. Immunogenicity studies have shown that antispike antibody levels whether elicited by vaccination or natural infection are a correlate of vaccine protection against COVID-19 disease [18]. Te incidence of breakthrough infections is found to increase with waning immunity in older adults and clinical high-risk groups [19,20] and with vaccine evasion by variant strains [21]. In a study by Gilbert et al. [22] evaluating SARS CoV 2 antibody markers as correlates of protection after mRNA-1273 vaccination against COVID-19, it was concluded that binding and neutralizing antibody markers strongly inversely correlated with COVID-19 risk and directly correlated with vaccine efcacy. Our study corroborates with this study that low S1/ S2 antibody titre was associated with an increased risk of mortality.
Mortality was also seen in 5vaccinated patients who had high S1S2 antibody titres (Table 4). Mortality in these was attributed to factors such as presence of multiple comorbidities, secondary infections, or receipt of high dose of steroids. In a study conducted in US by CDC, 1,228,664 individuals who had completed primary vaccination were studied [23]. In these individuals, 189 (0.015%) had severe COVID-19 outcomes, and 36 (0.0033%) individuals died. Tey also noted that all individuals with severe outcomes    were ≥ 65 years age, were immunosuppressed, or had at least one comorbidity. Our study has limitations because it is a single center retrospective study. A prospective study with serial antibody measurement for evaluating the vaccine efectiveness and waning immunity may ofer better insight. Another limitation is that the number of patients in unvaccinated group is higher. Tis is because in the initial phase, we saw the surge of delta virus afected cases who were largely unvaccinated cases. After the decline in the surge, we saw mixed vaccinated and unvaccinated cases in the ICU that is from mid-August onwards, that is, when we started measuring the antispike antibody levels as a marker of vaccine protection.

Conclusion
Our study lends support to the fnding that even in ICU setting, vaccination is associated with lower mortality. Hence, our fndings support the recommendation that antispike antibody levels be monitored for waning vaccine immunity. Future studies are needed to defne an objective level of antispike antibody titres at which booster dose may be recommended especially in elderly individuals with comorbidities.

Data Availability
Te data used to support the fndings of this study are included within the article.

Conflicts of Interest
Te authors declare that they have no conficts of interest.