Immune Thrombocytopenia (ITP) and Hβ-1 Tubulin: The Arg307 His Substitution Leads to Increased Treatment Failures and Increased Platelet Turnover

Abstract 525

Introduction. Hβ-1 tubulin is a β-tubulin isotype required for platelet production and discoid shape. It is the only tubulin isotype with non-synonymous single nucleotide polymorphisms (SNPs), some of which lead to changes in platelet number, shape, and function. We have previously reported that in ITP, the R307H substitution in Hβ-1 tubulin is associated with more severe thrombocytopenia and the use of significantly more treatment modality types during treatment of ITP. It was not known, however, if the effects of R307H in ITP occurs via alteration in platelet production and/or turnover, and whether treatment outcomes are affected by the SNP.

Methods. We sequenced exon 4 of TUBB1 (encoding Hβ-1 tubulin) using genomic DNA extracted from whole blood of 200 ITP patients to determine the presence of rs6070697 which leads to the the substitution of arginine for histidine at amino acid 307 (R307H). The first 100 patients in this series were the basis of our previous reports. Genotypes are reported as wild type (R/R), heterozygote (R/H), and homozygote SNP (H/H). A retrospective analysis of demographics and disease characteristics was performed for all patients. Patient responses to treatment modalities—classified as TPO-mimetic (TPO-m; eltrombopag, romiplostim, AKR-501), and non-TPO-mimetic treatments (non-TPO-m; all others)—were grouped by platelet counts (No Response (NR): 0–30; Partial Response (PR):31-100; Complete Response (CR):101-450). For a subset of patients, the absolute immature platelet fraction (A-IPF; immature platelet fraction (%) × platelet count) was also determined. The A-IPF is directly correlated with platelet production and turnover. The mean A-IPF for patients within each Hβ-1 genotype group was calculated and subsequently correlated with patient response to type of treatment. Fisher's exact test was used to compare platelets at initial presentation, number of treatment modalities, and response to treatments between genotypes; two-tailed Student's t-test was used to compare A-IPF measurements between genotypes.

Results. Baseline demographics and disease characteristics were not significantly different between the three Hβ-1 genotypes, including sex, race, age at ITP onset, median follow-up of ITP, or presence of other autoimmune diseases. We found that homozygote SNP patients (H/H) presented with more severe thrombocytopenia requiring immediate treatment (platelet <30 × 10⁹/L: H/H 100%, R/H 43%, R/R 64%, p=0.01). Median number of treatment modality types required was significantly higher for H/H patients compared to R/H and R/R (7.5, 5, and 4 treatment types, respectively; p=0.001). R/H and H/H patients had significantly higher (doubled) rate of treatment failures with non-TPO-m treatments compared to R/R patients (Failure rates 31% for R/H and H/H, 18% for R/R; Hazard Ratio 0.57; 95%CI 0.4–0.8; p=0.004). Furthermore, median A-IPFs were significantly higher in the R/H and H/H genotypes compared to the R/R group when a CR was achieved using non-TPO-m treatments (A-IPF=11.4 × 10⁹/L for R/H and H/H, 6.7 × 10⁹/L for R/R, p=0.005). In addition, the mean peak A-IPFs were significantly higher in the R/H and H/H genotypes compared to the R/R (20 × 10⁹/L for R/H and H/H, 10 × 10⁹/L for R/R, p=0.006). In contrast, when TPO-m were used we did not observe any significant difference between patient responses or A-IPFs across genotypes.

Conclusion. Hβ-1 tubulin R307H substitution was strongly associated with double the rate of treatment failures to non-TPO-m modalities. This contrasts with TPO-m treatments where no difference was found. Mechanistically, it appears that the R307H substitution leads to increased platelet turnover in ITP patients, as evidenced by a significantly higher immature platelet number at normal platelet counts in patients with the substitution when non-TPO-m modalities were used. TPO-m treatments appear to overcome the differences imparted by the substitution. The high prevalence of the R307H SNP in the general population (30% for R/H and H/H) together with the strong clinical association with treatment failure to non-TPO-m treatments in ITP treatment suggest that this Hβ-1 SNP may be a predictive biomarker to guide clinical decision making.