Low Thrombopoietin and Elevated Fas-L Are Independent Predictors of Immune Thrombocytopenia

Introduction. In immune thrombocytopenia (ITP), great progress has been made in recent years in terms of therapeutic options. In comparison, little has changed in the field of diagnostics. There are a number of studies on potential ITP biomarkers, but many of them are incomplete, unsystematic or were conducted without adequate controls. Thus, ITP remains a diagnosis of exclusion. In this study, we used a well-characterised cohort of patients with active ITP to investigate which of the published potential serum-based biomarkers could actually be suitable as predictor of the disease.

Methods. Adult patients with active ITP receiving no treatment were selected in accordance with current ASH ITP guidelines. In order to increase diagnostic specificity, only patients with a positive direct glycoprotein-specific test were included in the cohort (ITP, n=26). Patients with thrombocytopenia but no ITP were included in the low platelet control group (LPC, n=27), and healthy blood donors in the normal control group (NC, n=13). Following an extensive literature search for serum markers of ITP, we finally designed a customized panel for the detection of 22 cytokines based on Luminex multiplex technology (R&D Systems, Minneapolis, MN, USA). Serum samples stored at -80 °C were single-thawed and batch tested. Data were analyzed using SPSS (IBM, Armonk, NY, USA).

Results. Comparing the ITP group with the LPC group revealed a statistical significant difference for 9 out of 22 cytokines measured. Namely, IFNg, CXCL11, thrombopoietin (TPO), and TNFa concentrations were lower in ITP patients, whereas concentrations of CD30, CXCL11, Fas-L, IL2, IL17A, and IL18 were higher in ITP patients than in non-ITP patients with low platelet counts. In the univariate analysis, the following cytokines were significantly associated with a diagnosis of ITP: Fas-L (P<0.001, R²=0.80), TPO (P<0.001, R²=0.61), IL2 (P=0.05, R²=0.24) and CD40L (P=0.033, R²=0.18). However, in the multivariate analysis, only Fas-L and TPO remained independently predictive parameters for ITP (P<0.001, R²=0.92). The same model was also predictive when comparing ITP and NC (P<0.001, R²=0.87). The median Fas-L concentration in ITP sera was 6.5-fold higher than in the LPC cohort, and 3.6-fold higher than in normal controls. TPO concentration in ITP sera was only 1/10 of that in the LPC cohort and 1/5 of that in the NC cohort. Neither Fas-L nor TPO showed a correlation with platelet counts in any cohort.

Discussion. This, to our knowledge, is the first study on ITP patients which tested a large number of biomarkers in parallel. The strength of our study is the robust definition of ITP cases and the use of an adequate thrombopenic control group, which circumvents the influence of platelet mass on the concentration of individual biomarkers. Some of the 22 biomarkers reported in literature and investigated here were significantly different between ITPs and controls, but interestingly, the majority was not. The crucial finding is that multivariate analysis reduced their number to only two: TPO and Fas-L. TPO stimulates the proliferation, size and ploidy of megakaryocytes. Its use as a potential diagnostic marker has been discussed for more than 30 years after it was observed that unlike almost all other patients with low platelets, in whom the TPO concentration is elevated, low platelets in ITP patients do not trigger higher TPO levels. The mechanism behind this phenomenon is still largely unresolved. Fas-L is a membrane-associated cytokine, which can be shed from lymphocytes. Increased concentration in ITP sera may simply indicate T cell activation, as described for other autoimmune disorders such as, rheumatoid arthritis. However, soluble Fas-L acts through Fas receptor to stimulate the Caspase-8 dependent extrinsic pathway of programmed cell death. Megakaryocytes possess a Fas-L inducible apoptotic pathway, even though this pathway seems dispensable for platelet production, at least in mice.

Conclusion. In conclusion, our study has shown that the combination of two biomarkers in serum may detect more than 90% of cases with active ITP in patients with low platelet counts. Of course, this finding needs to be verified in a larger group, but it could represent an unexpectedly significant step forward in the diagnosis of ITP.