Platelet Laboratory Markers of Failure of First-Line Treatments in Immune Thrombocytopenia (ITP)

Introduction. Immune thrombocytopenia (ITP) is caused by a complex interplay between increased platelet destruction and/or impaired platelet production. First-line treatment includes corticosteroids (CS) and intravenous immunoglobulin (IVIG). Up to 15-20% of patients are refractory to first-line treatment and require further lines of treatment. Currently, however, there are no clinical or laboratory predictors of failure of first-line treatment(s). The aim of our study was to identify potential platelet laboratory markers linked to the failure of first-line treatment(s) in ITP.

Methods. All patients were identified from the Benign Hematology outpatient clinic of Ospedale San Paolo, Milan. We included adult patients with a diagnosis of ITP who had previously received CS and/or IVIG in accordance with clinical guidelines, with the latest course being the index. All patients underwent venous blood withdrawal at least 14 days after the start of CS and 3 days after the administration of IVIG. We evaluated reticulated platelets, markers of platelet activation [platelet-monocyte aggregates (PMAs) and platelet-granulocyte aggregates (PGAs)] and the degree of platelet desyalilation (assessed through platelet beta-galactose exposure) using flow cytometry. The cut-off for all evaluations was the 95^th centile of 20 healthy subjects. Patients were categorized in three groups based on the quality of the platelet count response at the time of blood sampling after the index course of first-line treatment: poor (0-49x10⁹/L), good (50-99x10⁹/L), and optimal (100x10⁹/L or more). Clinical records were screened up to 4 weeks post-sampling to assess first-line treatment failure (TF), which was defined as a platelet count below 25x10⁹/L, active bleeding or need for subsequent treatment.

Results. Twenty-eight patients were enrolled in this study. The median years since ITP diagnosis until study inclusion were 3 (IQR 1-13). Index first-line treatments included: CS (61% of patients), IVIG (7%) and CS+IVIG (32%). Nine patients had a poor response, 6 a good response, and 13 an optimal response to the index first-line treatment. In the poor response group, there was a lower proportion of females (44% vs 83% and 69% in the good and optimal response groups) and of secondary ITP (0% vs 17% and 31% in the good and optimal response groups) and the median age was higher (63 years, IQR 50-69, vs 40 years, IQR 37-55, and 55 years, IQR 41-65, in the good and optimal response groups). The proportion of patients with high reticulated platelets in the poor response group was 33% (3/9), which was non-significantly higher compared to the optimal response group (0/12, 0%, p=0.06) and the good response group (1/6, 17%, p=0.60). Similarly, the highest proportion of patients with elevated PMAs and PGAs was found in the poor response group (PMAs: 7/9, 78%; PGAs: 7/9, 78%) compared to the good response (PMAs: 2/6, 33%; PGAs: 3/6, 50%) and optimal response (PMAs, 3/13, 23%; PGAs, 5/13, 38%) groups. The proportion of ITP patients with increased PMAs was significantly higher in the poor response group compared to the optimal response group (p=0.03). Increased beta-galactose exposure was found in 50% of tested patients from the poor (3/6) and good (2/4) response groups, a proportion that was non-significantly higher compared to the optimal response group (2/10, 20%; p>0.10 for both comparisons). The 4-week follow-up post-blood sampling was complete for 25 patients. TF occurred in 5 patients, 3 in the poor response (3/8, 38%) and 2 (2/4, 50%) in the good response group. None of the TF patients had increased reticulated platelets (0/5, 0%). Increased PMAs and PGAs were found in 60% (3/5) of patients, and increased beta-galactose exposure in 67% of tested patients (2/3).

Conclusions. Our results suggest that the quality of platelet count response following first-line treatment parallels the downregulation of diverse platelet laboratory markers. A higher proportion of patients with a poor response had increased reticulated platelets, markers of platelet activation and platelet desyalilation, which suggest a higher degree of ongoing Fc-dependent and -independent platelet clearance. However, none of the investigated markers was able to discriminate patients who underwent TF at 4 weeks post-sampling. Whether an inappropriately low reticulated platelet count signals the need for second-line treatment needs further investigation.