Safety and Efficacy of Tyrosine Kinase Inhibitors in Immune Thrombocytopenic Purpura: A Systematic Review of Clinical Trials

Immune thrombocytopenic purpura (ITP) is an acquired antibody or cell-mediated platelet damage or decreased platelet production. Steroids, IV immunoglobulins (IVIG), and Rho-anti-D antibodies are the commonly used initial treatments for ITP. However, many ITP patients either do not respond or do not maintain a response to initial therapy. Splenectomy, rituximab, and thrombomimetics are the commonly used second-line treatment. More treatment options include tyrosine kinases inhibitors (TKI), including spleen tyrosine kinase (Syk) and Bruton’s tyrosine kinase (BTK) inhibitors. This review aims to assess the safety and efficacy of TKIs. Methods: Literature was searched on PubMed, Embase, WOS, and clinicaltrials.gov using keywords, “tyrosine kinase” and “idiopathic thrombocytopenic purpura”. PRISMA guidelines were followed. Results: In total, 4 clinical trials were included with 255 adult patients with relapsed/refractory ITP. In all, 101 (39.6%) patients were treated with fostamatinib, 60 (23%) patients with rilzabrutinib, and 34 (13%) with HMPL-523. Patients treated with fostamatinib achieved a stable response (SR) and overall response (OR) in 18/101 (17.8%) and 43/101 (42.5%) of the patients, respectively, while SR and OR were achieved in 1/49 (2%) and 7/49 (14%) of the patients, respectively, in the placebo group. Patients treated with HMPL-523 (300 mg dose expansion) achieved an SR and OR in 5/20 (25%) and 11/20 (55%) of the patients, respectively, while SR and OR were achieved in 1/11 (9%) of the patients treated with the placebo. Patients treated with rilzabrutinib achieved an SR in 17/60 (28%) patients. Dizziness (1%), hypertension (2%), diarrhea (1%), and neutropenia (1%) were serious adverse events in fostamatinib patients. Rilzabrutinib or HMPL-523 patients did not require a dose reduction due to drug-related adverse effects. Conclusions: Rilzabrutinib, fostamatinib, and HMPL-523 were safe and effective in the treatment of relapsed/refractory ITP.


Introduction
Immune thrombocytopenic purpura (ITP) is an acquired form of platelet cell destruction due to an antibody or cell-mediated platelet damage or impaired platelet production, kinase inhibitors (TKIs) in treating persistent ITP and a comparison with current treatment options. We will also discuss ongoing clinical trials and the need of future clinical trials.

Materials and Methods
Cochrane [13] and PRISMA [14] guidelines were followed by the authors in this systematic review.

Search Strategy
A comprehensive search was made on PubMed (Medline), Ovid Embase, Web of Science (WOS), and registry of clinicaltrials.gov with keywords, "tyrosine kinase" and "idiopathic thrombocytopenic purpura". The literature search was performed from the beginning of the data until 21 October 2022. The PICO framework was used to perform this literature search (Table S1) [15].

Inclusion and Exclusion of Articles
All the clinical trials providing safety (adverse effects) and efficacy (platelet response) data on TKIs inhibitors in AML were included. All the review articles, case reports, preclinical studies, and clinical studies irrelevant to TKI drugs or ITP were excluded. All the clinical trials without any safety or efficacy outcomes were also excluded.

Study Selection
Articles were screened by two authors (ZO and MH) and included based on pre-specified inclusion criteria. A third researcher (MAA) addressed the differences in screening.

Data Extraction
Two authors (GD and MYA) extracted the relevant data for the baseline characteristics of the included studies (treatment medication with dose, median age, splenectomy history, previous therapies, baseline platelet count), efficacy outcomes (stable response rate (SR), overall response (OR), modified stable response (MSR), and adverse events (≥grade 3 adverse effects). A stable response was defined as platelets ≥ 50,000/mm 3 ≥ 4 of the six visits biweekly. An overall response was ≥1 platelet count ≥ 50,000/mm 3 in the first 12 weeks of treatment. A modified stable response was defined as ≥2 platelet counts of ≥50,000/mm 3 , separated by a minimum of 5 days.
A primary outcome was a stable response. An overall response and a modified stable response were secondary outcomes for efficacy. An incidence of ≥grade 3 adverse effects were safety outcomes.

Risk of Bias (ROB) Assessment
ROB was conducted by using the Cochrane ROB-II tool by two researchers (WA and MZ) [16].

Risk of Bias
The ROB was high in Kuter et al. 2022 as it was a single-arm phase I/II study and lacks randomization and blinding. There was some concern of bias in Yang et al. 2021 as insufficient data were available regarding the allocation of treatment regimen for each patient and blinding. Treatment outcomes, stable and overall response were not provided in all the subgroups of patients. The ROB was low in Bussel

Risk of Bias:
The ROB was high in Kuter et al. 2022 as it was a single-arm pha lacks randomization and blinding. There was some concern of bias in Y insufficient data were available regarding the allocation of treatment patient and blinding. Treatment outcomes, stable and overall response w in all the subgroups of patients. The ROB was low in Bussel et al. 2018; F

Efficacy and Safety of TKI:
In 4 clinical trials, 255 adult patients were treated with TKIs, 101 (39.6%) patients with fostamatinib, 60 (23%) patients with rilzabrutinib, and 34 (13%) with HMPL-523. In all, 43 (17%) patients had a relapse despite splenectomy, no prior splenectomy in 167 (65.4%) patients, and unknown splenectomy status in 45 (17.6%) patients. All the patients had prior steroid treatment. Baseline characteristics of the patients are given in; Table 1.    Concomitant = Multiple ITP mechanisms involved (i.e., ITP destruction and decreased production), Persis-tent=between 3 to 12 months from diagnosis not reaching spontaneous remission or not maintaining complete response off therapy, Chronic=lasting for more than 12 months.

Efficacy of Fostamatinib and HMPL-523 (Syk)
In two phase III RCTs (N = 135) by Bussel et al. 2018, patients were previously treated with a median of three lines of treatment. Patients in the fostamatinib group achieved an OR and SR in 43/101 (42.5%) and 18/101 (17.8%) of the patients, respectively, while OR and SR were achieved in 7/49 (14%) and 1/49 (2%) of the patients, respectively, in the placebo group. In a phase Ib RCT study by Yang et al. 2021 (N = 60), patients treated with HMPL-523 (300 mg dose expansion) achieved an SR and OR in 5/20 (25%) and 11/20 (55%) of the patients, respectively, while OR and SR were achieved in 1/11 (9%) of the patients in the placebo group; Table 2.

Efficacy of Bruton's TKI (Rilzabrutinib)
In the phase I-II clinical trial by Kuter et al. (N = 60), patients were previously treated with a median of four lines of treatment. Patients treated with rilzabrutinib achieved an SR and MSR in 17/60 (28%) and 24/60 (40%) patients, respectively. The median time to platelet count of >50 × 10 3 /mm 3 was 11.5 days; Table 2.

Safety
On treatment with fostamatinib, dizziness, hypertension, neutropenia, and diarrhea were serious adverse events reported in 1%, 2%, 1%, and 1% of the patients, respectively. Dose reduction due to adverse events was reported in 9% of the patients treated with fostamatinib. On treatment with rilzabrutinib, grade 1/2 diarrhea, nausea, fatigue, and vomiting were reported in 37%, 35%, 20%, and 7% of the patients, respectively. No treatment-related grade 3 or 4 adverse effects were seen on patients on rilzabrutinib. On treatment with HMPL-523, no adverse effects led to a dose reduction. Low-grade increased ALT, LDH, bilirubin, lipids, and blood pressure were reported in 25%, 25%, 20%, 10%, and 10% of the patients, respectively, on treatment with 300 mg HMPL-523; Table 2.

Efficacy of Bruton's TKI (Rilzabrutinib)
In the phase I-II clinical trial by Kuter et al. (N = 60), patients were previously treated with a median of four lines of treatment. Patients treated with rilzabrutinib achieved an SR and MSR in 17/60 (28%) and 24/60 (40%) patients, respectively. The median time to platelet count of >50 × 10 3 /mm 3 was 11.5 days; Table 2.

Safety
On treatment with fostamatinib, dizziness, hypertension, neutropenia, and diarrhea were serious adverse events reported in 1%, 2%, 1%, and 1% of the patients, respectively. Dose reduction due to adverse events was reported in 9% of the patients treated with fostamatinib. On treatment with rilzabrutinib, grade 1/2 diarrhea, nausea, fatigue, and vomiting were reported in 37%, 35%, 20%, and 7% of the patients, respectively. No treatment-related grade 3 or 4 adverse effects were seen on patients on rilzabrutinib. On treatment with HMPL-523, no adverse effects led to a dose reduction. Low-grade increased ALT, LDH, bilirubin, lipids, and blood pressure were reported in 25%, 25%, 20%, 10%, and 10% of the patients, respectively, on treatment with 300 mg HMPL-523; Table 2.

Safety
On treatment with fostamatinib, dizziness, hypertension, neutropeni were serious adverse events reported in 1%, 2%, 1%, and 1% of the patien Dose reduction due to adverse events was reported in 9% of the patien fostamatinib. On treatment with rilzabrutinib, grade 1/2 diarrhea, nause vomiting were reported in 37%, 35%, 20%, and 7% of the patients, respect ment-related grade 3 or 4 adverse effects were seen on patients on rilzabru ment with HMPL-523, no adverse effects led to a dose reduction. Low-g ALT, LDH, bilirubin, lipids, and blood pressure were reported in 25%, and 10% of the patients, respectively, on treatment with 300 mg HMPL-52

Discussion
Fostamatinib is an orally bioavailable competitive inhibitor of the Syk catalytic domain, tested in the treatment of ITP in adults with no or inadequate response to a prior therapy. Syk is expressed on B cells, macrophages, T cells, and platelets. Syk is activated when the Fc gamma receptor binds to its ligand and leads to the phosphorylation of activation tyrosinebased motifs in the immunoreceptor. In macrophages, these motifs lead to cytoskeletal changes and phagocytosis of platelets, while in B cells, Syk may have a role in antibody formation [20].
Intestinal alkaline phosphatase converts fostamatinib to an active metabolite R406. The active metabolite can inhibit the Fc epsilon receptor RI and Fc gamma receptor on mast cells, as well as Syk-dependent signaling on B cells [21]. Other possible mechanisms may include the inhibition of Jak, Lck, and Flt-3 pathways [21,22]. FIT1 and FIT2 were the multicenter phase III RCTs conducted by authors in Bussel et al. 2018 [17]. Fostamatinib was able to produce a response in a significant number of the patients who did not respond to TPO therapy, rituximab, and/or splenectomy. On further analysis by authors in Bussel et al. 2018, the response rate with fostamatinib was relatively higher in younger patients and platelet counts of 15-30 × 10 3 /mm 3 as compared to older patients and platelet counts of <15 × 10 3 /mm 3 . The response rate in patients with antiplatelet antibodies had a higher response rate as compared to patients without detectable antibodies (36% vs. 9%). Additionally or within the first 12 weeks were higher in patients with fostamatinib as the second line of therapy, as compared to the third, fourth, or fifth lines of therapy (78%, 64%, 52%, 36%, respectively).
Common treatment-related adverse events were mild/moderate and included hypertension, diarrhea, nausea, and transaminase elevation. No bleeding-related serious adverse events were reported. Medical management with antimotility and antihypertensives were sufficient for the patients with hypertension and gastrointestinal adverse effects. In the long-term phase III 5-year extension study by Cooper et al. 2021 (N = 146) [23], no new toxicities were observed. Only one thromboembolic event was reported as a mild transient ischemic attack that is much less than expected in an ITP patient. Platelet counts of >50 × 10 3 /mm 3 were reported in 54% of the patients. Initial results of a real-world study of fostamatinib were presented in the annual meeting of ASH 2022 conducted by authors in Moezi et al. 2022 (N = 46) [24]. The median number of prior therapies was 2 in this study. Platelet counts of >30 × 10 3 /mm 3 and >50 × 10 3 /mm 3 were achieved by 31 (67.4%) and 26 (56.5%) of the patients, respectively.
Fostamatinib is available in the market at a cost of about 11,000 USD for 1 month, which is comparable to the price of other oral agents such as eltrombopag (around 10,000 USD) and injectable agent romiplostim (around 9000 USD). Therefore, patient preference and medical necessity are major determinants of which agent to use rather than cost-effectiveness [25].
HMPL-523 is an oral agent that reversibly inhibits Syk; thus, preventing the destruction of opsonized platelets and auto-antibody formation, which is still under development and just completed its Phase 1b trial in China (NCT03951623) [18]. In this trial, HMPL-253 was well tolerated with no serious adverse events at a dose of 400 mg daily with a documented efficacy against the placebo. The adverse events seen include derangements in the liver-function test, hyperlipidemia, elevated amylase levels, hypokalemia, dizziness, diarrhea, proteinuria, and hypertension. In another study conducted in healthy Australian males, HMPL-523 caused febrile illness and elevated lipase levels [26]. HMPL-523 needs a high-powered study and head-to-head trials with other treatments such as rituximab or splenectomy to establish its relative efficacy. A phase III trial NCT05029635 is in progress to further assess the safety and efficacy of HMPL-523 in patients with refractory ITP. It can be used as an alternative agent to fostamatinib for SYK inhibition [27,28]. The drug is not yet approved by FDA and is not available in the market for ITP.
Rilzabrutinib is an oral BTK inhibitor (BTKI) that plays an important role in B-cell function, maturation, differentiation, and antibody formation [29]. Among other pathways, BTK is involved in Fc gamma receptor signaling and involves decreased macrophagemediated platelet destruction and auto-antibodies formation. It has covalent and noncovalent binding sites; therefore, it can bind with high potency and can have long binding time [29]. Rilzabrutinib has a rapid on rate and a slow off rate with >80% binding in an hour and maintains receptor occupancy for 24 h. Unbound rilzabrutinib rapidly clears out of the system within 6 h, limiting systemic toxicity. In preclinical studies, rilzabrutinib, unlike other BTK inhibitors ibrutinib, does not affect the PI3K-Akt pathway that was believed to be associated with adverse effects such as atrial arrhythmia [29]. Similarly, older generation BTKIs such as ibrutinib also inhibit other kinases, which leads to decreased collagen-mediated platelet aggregation and are associated with bleeding diatheses. No such adverse effects were seen with rilzabrutinib in the short-term follow up [30]. The phase I/II clinical trial (NCT03395210) by Kuter et al. 2022 [19] included highly refractory ITP patients with a median of four lines of prior treatments (glucocorticoid, rituximab, TPO, IVIG, fostamatinib, splenectomy). In total, 400 mg twice daily was the highest dose tested in the trial and was able to produce a response in 40% of the patients; in addition, it was well tolerated by most of the patients. Further trials and testing are needed for the long-term efficacy and durability of rilzabrutinib treatment in patients with relapsing and refractory immune thrombocytopenia. Rilzabrutinib is under consideration by FDA for approval based on these results. A pivotal LUNA 3 phase III RCT (NCT04562766) is underway to assess rilzabrutinib treatment efficacy and safety. Therefore, it is not available in the market for ITP; however, it can be found for other diseases such as pemphigus vulgaris and may become available soon after FDA approval.
Trials are in progress for newer agents targeting tyrosine kinases, such as baritinib, orelabrutinib, zanubrutinib, and SKI-O-703. Baricitinib is an oral agent that binds reversibly with JAK proteins leading to the inhibition of the JAK-STAT pathway involved in gene transcription and the downstream activation of inflammatory mediators involved in autoimmune responses [31]. An early-phase trial is in progress in China to assess the safety and efficacy in steroid relapsed/refractory ITP patients (NCT05446831). Orelabrutinib is a BTKI-like rilzabrutinib and is involved in multiple signaling pathways of adaptive or innate immunity. In a preclinical study by Yu et al. 2021, the expression of CD69 and the BCR signaling pathway CD86 was significantly reduced with orelabrutinib in polymorphonuclear cells separated from ITP and healthy human subjects. In mice, the administration of orelabrutinib significantly increased the platelet count. Therefore, it can be an option for ITP patients in the future and human trials are in progress in China (NCT05124028, NCT05020288, and NCT05232149). Zanubrutinib is also a BTKI-like rilzabrutinib and orelabrutinib. A case report showed encouraging results of zanubrutinib in a 15-year-old Chinese girl with severe unresponsive ITP with Evans' syndrome [32]. Trials are in progress in China on monotherapy, and the combination of zanubrutinib with TPO and steroids in refractory ITP patients or first-line therapy (NCT05279872, NCT05369377, NCT05369364). SKI-O-703 is an Syk inhibitor such as fostamatinib and a phase II RCT is in progress in refractory ITP patients (NCT04056195).
Commonly used second line treatments for ITP are TPO, rituximab, and splenectomy. Splenectomy increases the infection risk especially with encapsulated bacteria, with reports of sepsis in 2-7 per person-year in splenectomy patients as well as an increased risk of thromboembolism. Surgical complications and the possibility of spontaneous remission of ITP in the first year, splenectomy is only considered after one year of diagnosis of ITP [33]. TKIs can be a valuable alternative for these patients for early use given a lesser number of severe adverse effects and may delay or completely obliviate the need of a splenectomy in these patients. Similarly, the combination of TKIs with a splenectomy can also be tested in clinical trials to reduce the risk of relapse since TKIs were able to produce a response in patients relapsed/refractory to splenectomy.
Rituximab is an off-label treatment of ITP for years. An early trial on rituximab in combination with dexamethasone as a first-line therapy showed encouraging results [34]; or in the range of 40-70%, were reported in patients after 4 weeks of rituximab of 375 mg/m 2 and on 5 years of follow up; the response was sustained in about 21% of the patients [10,35,36]. In the long-term RCT study by Ghanima et al. 2015 [10], there was no significant long-term benefit reported with the use of rituximab as second line of therapy. Rituximab trials included patients who had a first line of treatment only as compared to trials on ITP that included patients who failed multiple lines of treatment. There is no direct study comparing rituximab with TKIs. An indirect comparison was performed by network meta-analysis on RCTs by authors in Laws et al. 2022 [37]. Fostamatinib was significantly more effective in the overall improvement of platelet count than rituximab regimens with an odds ratio of 0.11-0.2. The risk of infection and reactivation of hepatitis B infection with rituximab is also increased. Given the COVID-19 pandemic, the increased risk of infection and possible impairment of vaccine response in about the 6-month period after administration do not make rituximab an ideal choice in the current era [38].
TPO mimics endogenous TPO and improves platelet production by stimulating megakaryocyte maturation. Multiple adverse events, including bone marrow fibrosis, continuous stimulation of megakaryocytes, increased risk of thromboembolism, transaminitis, and severe rebound thrombocytopenia are associated with TPO. Along with the use of TKIs in patients relapsed/refractory to TPO therapy, TKIs can also be considered for use with TPO therapy. Concomitant use of TKIs with TPO may increase the response rate, the success rate of tapering of TPO by preventing rebound thrombocytopenia, and may decrease the risk of thromboembolic events [23]. There is an ongoing trial in China on the combination of zanubrutinib with TPO; however, more large-scale multicenter RCTs are needed to provide reliable evidence.