Effects of serum vitamin d on platelet function in primary immune thrombocytopenia Free

Background Vitamin D (VD) deficiency (VDD) has been associated with autoimmune diseases, though its exact role is unclear. Consistent with other autoimmune diseases, VDD has been linked to incidence and severity of primary immune thrombocytopenia (ITP), a bleeding disorder characterized by destruction and impaired production of platelets (PLT), mediated by autoantibodies and autoreactive T cells. Severe bleeding can occur in ITP, though bleeding symptoms tend to be mild relative to PLT counts. This is often attributed to immature, highly active PLT. Interestingly, ITP also exhibits higher risk of thrombosis. Recently, VDD was found to enhance PLT activation via the VDR/AKT pathway. Increased platelet reactivity may underlie the moderate bleeding severity and increased thrombotic risk in ITP. We aimed to assess the impact of VD levels on PLT function and investigate if VDD enhances PLT reactivity in ITP.

Methods Adult primary ITP patients from the Vienna ITP biobank were analysed (EC No.1843/2016). The SMOG score assessed bleeding severity.

Chemiluminescence immunoassay measured serum 25-OH-VD; levels ≥75nmol/L were sufficient (VDS), <75nmol/L and <50nmol/L were considered moderate VDD (mVDD) and severe VDD (sVDD). PLT function was determined by the expression of activated glycoprotein (GP) IIb/IIIa and P-selectin, measured by flow cytometry, naïve and after stimulation by agonists (5μg Thrombin Receptor Activator Peptide 6 (TRAP-6), 1μM adenosine diphosphate (ADP), 80μM arachidonic acid (AA), 0.04μg/mL collagen related protein (CRP)-XL). PLT reactivity and degranulation were defined as the change in GPIIb/IIIa or P-selectin expression respectively after stimulation.

Data was compared across VDD groups (sVDD, mVDD, VDS) with Kruskal-Wallis and Dunn's tests. PLT function results were associated with VD levels, sex, age, disease duration, platelet count, and cumulative SMOG score via multivariable linear regressions. Non-linear VD effects were explored using natural splines (5 degrees of freedom). Disease duration, PLT count, and cumulative SMOG score were log2-transformed. Significance of VD was determined by ANOVA F-test, comparing full and restricted models.

Results 79 patients (50 (63%) women; 29 (37%) men) with a median age of 47 years were analysed. 14 (17.7%) had sufficient VD, 32 (40.5%) had mVDD, and 33 (41.7%) had sVDD. No significant differences in median past/current treatment lines (sVDD: 2, mVDD: 2, VDS: 2) or steroid use at time of sample collection (7 (21%) sVDD, 6 (19%) mVDD, 5 (36%) VDS) were observed between groups.

PLT counts were lowest in sVDD patients (46x10⁹/L), lower than in VDS (53x10⁹/L) and significantly lower than in mVDD patients (78 x10⁹/L; p=0.012). sVDD had a significantly higher cumulative SMOG score than mVDD (p=0.016), though after adjusting for PLT count in ordinal regression, no relationship between bleeding severity and VD levels was found.

Naïve P-selectin and GPIIb/IIIa expression did not differ by VD status. While the increase of P-selectin expression in response to AA was higher in mVDD than sVDD and VDS (8% vs 2% vs 1%; p=0.006), there was no clear pattern of altered PLT response to stimuli by VD status. In multivariable regressions, naïve P-selectin or GPIIb/IIIa expression were not associated with VD levels, sex, age, ITP disease duration, PLT count, or bleeding severity. PLT degranulation and reactivity upon addition of agonists was significantly associated with higher PLT counts. No significant associations between VD levels and PLT function upon stimulation were found, except a non-linear association after CRP-XL addition, determined by increases in P-selectin and GPIIb/IIIa expression, with highest/lowest response at VD levels of 80/40nmol/L respectively.

Conclusion In our ITP cohort, we could not confirm an association of PLT function with VD levels or states of VD deficiency. Although differences in bleeding severity were observed between patients stratified by VD status, this did not coincide with clear differences in PLT function. While enhanced PLT reactivity was significantly associated with higher PLT counts, there were no clear relationships with VD levels in multivariable regression.

Though an association between VDD and ITP disease severity is plausible, the previous observation of increased platelet reactivity in VD deficiency could not be confirmed in adult primary ITP patients.