Publications

Long COVID (LC) involves a spectrum of chronic symptoms after acute severe acute respiratory syndrome coronavirus 2 infection. Current hypotheses for the pathogenesis of LC include persistent virus, tissue damage, autoimmunity, endocrine insufficiency, immune dysfunction and complement activation. We performed immunological, virological, transcriptomic and proteomic analyses from a cohort of 142 individuals between 2020 and 2021, including uninfected controls (n = 35), acutely infected individuals (n = 54), convalescent controls (n = 24) and patients with LC (n = 28). The LC group was characterized by persistent immune activation and proinflammatory responses for more than 180 days after initial infection compared with convalescent controls, including upregulation of JAK-STAT, interleukin-6, complement, metabolism and T cell exhaustion pathways. Similar findings were observed in a second cohort enrolled between 2023 and 2024, including convalescent controls (n = 20) and patients with LC (n = 18). These data suggest that LC is characterized by persistent activation of chronic inflammatory pathways, suggesting new therapeutic targets and potential biomarkers of disease.

IMPORTANCE: The US Preventive Services Task Force (USPSTF) guidelines on preeclampsia risk assessment and aspirin prophylaxis (AP) have not been evaluated for clinical utility.

OBJECTIVE: To evaluate which characteristics in the USPSTF guidelines identify risk status and the association of preeclampsia risk with AP recommendations.

DESIGN, SETTING, AND PARTICIPANTS: This observational cohort study enrolled from July 2020 to March 2023 with data analysis performed from October to December 2024. Enrollment occurred at 11 centers throughout the US or via direct-to-participant recruitment. Pregnant participants aged 18 years or older with a singleton pregnancy less than 22 weeks' gestation were selected via convenience sampling.

EXPOSURE: The exposures were clinical factors abstracted from medical records by research coordinators, which were stratified according to USPSTF definitions of low, moderate (parity, advanced maternal age [AMA], race, and body mass index), and high (chronic hypertension, prior preeclampsia, type 1 or 2 diabetes, kidney disease, and/or autoimmune conditions) risk.

MAIN OUTCOMES AND MEASURES: Data collected included AP recommendation, presence of USPSTF-defined moderate risk factors or high risk factors, and any preeclampsia diagnosis. Effect sizes and relative risk (RR) were calculated within risk strata.

RESULTS: Of 5684 participants (median [IQR] age, 30.9 [26.4-34.6] years; 267 [4.1%] Asian; 1191 [21.0%] Black; 990 Hispanic [17.4%]; 2764 [48.6%] White; and 472 [8.3%] with other race or ethnicity), 5046 (88.8%) were at increased risk of preeclampsia (3996 [70.3%] at moderate risk and 1050 [18.5%] at high risk). A total of 2438 participants (43.1%) received an AP recommendation. The overall preeclampsia rate was 12.1% (685 participants). The PE rates specific to USPSTF categories were 3.0% for those at low risk (19 of 638 participants), 10.5% for those at moderate risk (419 of 3996 patients), and 23.5% for those at high risk (247 of 1050 participants). Among individuals with 2 or more moderate risk factors but without any high risk factor, nulliparity was associated with significantly increased risk of preeclampsia (RR, 1.48; 95% CI, 1.35-1.62; P < .001), while AMA was associated with decreased risk (RR, 0.79; 95% CI, 0.65-0.96; P = .02); there was no association with obesity (RR, 1.11; 95% CI, 1.01-1.22; P = .048) or Black race (RR, 0.95; 95% CI, 0.80-1.14; P = .63). Of 1044 participants with any high risk factors, 856 (82.0%) were recommended AP and of 634 at low risk, 538 (85.9%) were not recommended AP. In contrast, of 1942 participants with 1 moderate risk factor, 463 (23.8%) were recommended AP, and of 2032 with 2 or more moderate risk factors, 1024 (50.4%) were recommended AP.

CONCLUSIONS AND RELEVANCE: In this prospective cohort study of 5684 singleton pregnancies, 89% of the population was assessed as having increased risk (moderate or high) of preeclampsia by USPSTF criteria. These findings suggest that moderate risk factors in the absence of high risk factors show no or low value for estimating the risk of developing preeclampsia, leading to nonspecific recommendations of AP in the moderate risk category.

Messenger RNA (mRNA) vaccines were highly effective against the ancestral SARS-CoV-2 strain, but the efficacy of bivalent mRNA boosters against XBB variants was substantially lower. Here, we show limited durability of neutralizing antibody (NAb) responses against XBB variants and isotype switching to immunoglobulin G4 (IgG4) responses following bivalent mRNA boosting. Bivalent mRNA boosting elicited modest XBB.1-, XBB.1.5-, and XBB.1.16-specific NAbs that waned rapidly within 3 months. In contrast, bivalent mRNA boosting induced more robust and sustained NAbs against the ancestral WA1/2020 strain, suggesting immune imprinting. Following bivalent mRNA boosting, serum antibody responses were primarily IgG2 and IgG4 responses with poor Fc functional activity. In contrast, a third monovalent mRNA immunization boosted all isotypes including IgG1 and IgG3 with robust Fc functional activity. These data show substantial immune imprinting for the ancestral spike and isotype switching to IgG4 responses following bivalent mRNA boosting, with important implications for future booster designs and boosting strategies.

The COVID-19 Biorepository at Beth Israel Deaconess Medical Center in Boston was initiated in 2020 to address questions about COVID-19 infection and vaccination in a time of urgent need. From April 2020 through July 2024, we enrolled 1018 participants and collected thousands of biospecimens. We enrolled participants from the general population as well as from specific populations that were not well represented in clinical trials, including immunosuppressed, pregnant, and lactating individuals. Our observational study was designed to accommodate the rapidly changing landscape of the pandemic, including the introduction of new vaccines and boosters, breakthrough infections, and emerging variants. Reflecting on the past four years of this experience, we believe that teamwork, collaboration, and flexibility were key factors for the success of this effort, which generated data in real time about COVID-19 vaccine responses in multiple populations, hybrid immunity following breakthrough infections, immune evasion of emerging variants, and immune imprinting following booster immunizations. Rapid dissemination of data through preprints, peer-reviewed publications, and public communications allowed for the real time use of our findings to address public health issues and to inform vaccine policies. The dedication of the study participants, clinical investigators, and laboratory investigators made this research program possible.

A new highly mutated Omicron subvariant BA.2.87.1 has recently been identified with over 30 amino acid mutations in the Spike protein compared with BA.2, BA.5, XBB.1.5, and JN.1 variants. Mutiple mutations in BA.2.87.1 are located in the N-terminal domain (NTD) rather than in the receptor binding domain (RBD) of the Spike protein. We evaluated neutralizing antibody (NAb) responses to BA.2.87.1 because of its highly mutated sequence and its unique NTD region. Our data show that NAb responses to BA.2.87.1 were lower than to BA.2 but higher than to JN.1, suggesting that BA.2.87.1 is not a further antibody escape variant compared with other currently circulating variants. Moreover, XBB.1.5 mRNA boosting increased NAb titers to all variants tested including BA.2.87.1.

Prior infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is typically measured by nucleocapsid serology assays. In this study, we show that the Simoa serology assay and T-cell intracellular cytokine staining assay are more sensitive than the clinical Elecsys assay for detection of nucleocapsid-specific immune responses. These data suggest that the prevalence of prior SARS-CoV-2 infection in the population may be higher than currently appreciated.

Current COVID-19 vaccines provide robust protection against severe disease but minimal protection against acquisition of infection. Intramuscularly administered COVID-19 vaccines induce robust serum neutralizing antibodies (NAbs), but their ability to boost mucosal immune responses remains to be determined. In this study, we show that the XBB.1.5 messenger RNA (mRNA) boosters result in increased serum neutralization to multiple severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants in humans, including the dominant circulating variant JN.1. In contrast, we found that the XBB.1.5 mRNA booster did not augment mucosal NAbs or mucosal IgA responses, although acute SARS-CoV-2 XBB infection substantially increased mucosal antibody responses. These data demonstrate that current XBB.1.5 mRNA boosters substantially enhance peripheral antibody responses but do not robustly increase mucosal antibody responses. Our data highlight a separation between the peripheral and mucosal immune systems in humans and emphasize the importance of developing next-generation vaccines to augment mucosal immunity to protect against respiratory virus infections.

The replication-incompetent modified vaccinia Ankara-Bavarian Nordic vaccine (MVA-BN; Jynneos) was deployed during the 2022 clade IIb mpox outbreak. On August 14, 2024, the World Health Organization declared the mpox clade Ib outbreak in the Democratic Republic of the Congo a public health emergency of international concern, which has raised the question about the durability of vaccine immunity after MVA-BN vaccination. In this study, we show that the MVA-BN vaccine generated mpox serum antibody responses that largely waned after 6-12 months.