ABSTRACT
Introduction: Cases of de novo immune thrombocytopenia (ITP), including a fatality following SARS-CoV-2 vaccination in a previously healthy recipient, led to studying its impact in pre-existing ITP. Published reports are limited but suggest that most patients with ITP tolerate the COVID-19 vaccines well without frequent ITP exacerbations (Kuter, BJH, 2021). Data regarding risk factors for exacerbation and relationship of response to first dose to that of second dose are limited. Methods: Data for patients with pre-existing ITP were obtained via 3 sources. First, via a ten-center retrospective study of adults with ITP who received a SARS-CoV-2 vaccine between December 2020 and March 2021 and had a post-vaccination platelet count (n=117);9 centers were in the United States. Eighty-nine percent of patients received mRNA-based vaccines. The second and third sources of data were surveys distributed by the Platelet Disorder Support Association (PDSA) and the United Kingdom ITP Support Association. A 'stable platelet count' was defined as a post-vaccination platelet count within 20% of the pre-vaccination level. ITP exacerbation was defined as any one or more of: platelet decrease ≥ 50% compared to pre-vaccination baseline, platelet decrease by >20% compared to prevaccination baseline with platelet nadir < 30x10 9/L, receipt of rescue therapy for ITP. Continuous variables were described as mean ±SD or median [interquartile range];categorical variables were described as n (%). Relative risks and 95% confidence interval were calculated to estimate strength of association. Results: Among 117 patients with pre-existing ITP from 10 centers who received a SARS-CoV-2 vaccine, mean age was 63±17 years, 62% were female, with median 12 [4-23] years since diagnosis of ITP;patients had received a median of 3 [2-4] prior medical treatments. Sixtynine patients were on ITP treatment at the time of vaccination (Table 1). There was an almost even distribution of platelet count response following each vaccine dose. In 109 patients with data for dose 1, platelet counts increased in 32 (29%), were stable in 43 (39%), and decreased in 34 (31%);in 70 patients following dose 2, platelet counts increased in 24 (34%), were stable in 25 (36%), and decreased in 21 (30%) (Figure 1). Nineteen (17%) patients experienced an ITP exacerbation following the first dose and 14 (20%) of 70 after a second dose. In total, fifteen patients received and responded to rescue treatments (n = 6 after dose 1, n = 8 after dose 2, n = 1 after both doses). Of 7 patients who received rescue treatment after dose 1, 5 received dose 2 and only 1/5 received rescue treatment again. Rescue consisted of increased dose of ongoing medication, steroids, IVIG, and rituximab. Splenectomized persons and those who received 5 or more prior lines of medical therapy were at highest risk of ITP exacerbation. Only 1 of 47 patients who had neither undergone splenectomy nor received 5 or more lines of therapy developed ITP exacerbation after dose 1. There were 14 patients offtreatment at the time of dose 1 and 7 patients at time of dose 2;1 patient in each group developed ITP exacerbation with both these having had normal counts prior to vaccination and having undergone splenectomy. In 43 patients whose platelet counts were stable or increased after dose 1 and received dose 2, only 6 (14%) had platelet decreases to <50 x10 9/L after dose 2. Age, gender, vaccine type, and concurrent autoimmune disease did not impact post-vaccine platelet counts. In surveys of 57 PDSA and 43 U.K. ITP patients, similar rates of platelet change were seen (33% of participants reported decreased platelet count in both surveys) and prior splenectomy was significantly associated with worsened thrombocytopenia in each. Conclusions: Thrombocytopenia may worsen in pre-existing ITP post-SARS-CoV2-vaccination but when ITP exacerbation occurred, it responded well to rescue treatment. No serious bleeding events were noted. Rescue treatment was needed in 13% of patients. Proactive vaccination surveillance of patien s with known ITP, especially those post-splenectomy and with more refractory disease, is indicated. These findings should encourage patients with ITP to not only be vaccinated, but to receive the second dose when applicable to ensure optimal immunization. Rituximab interferes with vaccination response and ideally would be held until a minimum of 2 weeks after completion of vaccination.
ABSTRACT
VITT is an immune-based complication of adenoviral-based vaccines used to immunize against SARS_CoV2. The antibodies in VITT have been described as directed at the platelet-specific chemokine PF4 (CXCL4). While the clinical course and target chemokine in VITT has much in common with the better-known thrombocytopenic/prothrombotic disorder, heparin-induced thrombocytopenia (HIT), which involves antibodies directed against PF4 bound to the polyanion heparin, the specific loci where VITT and PF4/polyanion HIT antibodies bind appear to differ in studies using alanine-scanning mutations of PF4 (Nature, 2021. DOI: 10.1038/s41586-021-03744-4). The VITT antigenic site localizes to a heparin-binding domain. Unlike the dominant HIT locus, the VITT locus is conserved not only between human and mouse PF4, but also between PF4 and the related platelet-specific chemokine NAP2 (CXCL7). NAP2 is also expressed and stored in platelet alpha-granules and is present in equimolar concentrations to PF4. Unlike PF4, NAP2 avidly binds the chemokine receptor CXCR2 and strongly activates neutrophils. We now show that antibodies from patients who developed VITT after both AstraZeneca (AZ) or Johnson and Johnson (JJ) adenoviral vaccines, unlike HIT antibodies, recognize mouse PF4 (Figure 1A). More importantly, both AZ and JJ VITT antibodies bound NAP2, while none of the HIT antibodies tested bound PF4 or NAP2 in the absence of heparin (Figure 1A). These results are consistent with the alanine-scanning studies that distinguish the HIT and VITT binding sites. Dynamic light scattering (DLS) showed that NAP2 and PF4 bind to the adenoviral vectors, including Ad5 and the AZ vector ChAdOx5, which leads to expression of SARS_CoV2 spike protein. ChAdOx2 vaccine and CsCl 2-purified ChAdOx2 bound to both proteins, but form larger complexes with NAP2 than with PF4 even at lower concentrations of this chemokine (Figure 1C). Removal of anti-PF4 antibodies by hPF4-Sepharose abrogated PF4-dependent binding, but did not significantly reduce binding to NAP2 (not shown), indicating that VITT plasma contains discrete pools of anti-PF4 and anti-NAP2 antibodies that may have distinct functional properties. Sandwich ELISA (not shown) and Western blot analysis of purified VITT IgG demonstrates the presence of hPF4-IgG and NAP2-IgG immune complexes in purified patient's IgG (Figure 2A). Functional studies show that both PF4 and NAP2 can activate platelets in the presence of VITT antibodies. Anti-PF4-depleted VITT IgG fraction retains the ability to activate platelets in the presence of NAP2 (Figure 2B). Thus, unlike HIT, VITT appears to target a shared antigenic site on the related chemokines PF4 and NAP2. This raises the question as to whether NAP2, as one the most abundant platelet chemokines released from activated platelets, is involved in the initiation and propagation of the immunothrombotic response. Additional studies are needed to see whether NAP2, which can potently and specifically activate neutrophils via CXCLR2, contributes to the specific thromboinflammatory phenotype seen in VITT. We propose using FcgammaRIIA+ mice that concurrently express human PF4 and NAP2 and specific knockout of each chemokine, available in our group, to further understand the pathogenesis of VITT and its thrombocytopenic/ prothrombotic phenotype. [Formula presented] Disclosures: Padmanabhan: Veralox Therapeutics: Membership on an entity's Board of Directors or advisory committees. Cines: Dova: Consultancy;Rigel: Consultancy;Treeline: Consultancy;Arch Oncol: Consultancy;Jannsen: Consultancy;Taventa: Consultancy;Principia: Other: Data Safety Monitoring Board.