Limitations of rituximab in treating primary immune thrombocytopenia and the therapeutic potentials of alternatives

Abstract

Purpose of review

Primary immune thrombocytopenia (ITP), an immune-mediated hemorrhagic disease, features an intricate pathogenesis that involves megakaryocyte malfunction and hyperresponsiveness of the innate and adaptive immune systems. As a second-line drug for ITP, rituximab acts quickly and can produce an initial response rate of up to 60%. However, this response only lasts for a short term, meanwhile challenged by resistance, relapse and side effects. Additionally, no reliable clinical parameters have been proposed for forecasting the therapeutic response of patients. Furthermore, the application of rituximab is restricted in specific populations, including pregnant patients, children with positive antithyroid antibodies, and patients contaminated with HBV.

Recent findings

Splenectomy and new drugs that target the thrombopoietin receptor, FcγR, FcRn, B-cells or plasma cells, T-cells, and complement pathways may overcome these shortcomings.

Summary

This article summarizes the barriers limiting the use of rituximab, and discusses the effectiveness and safety of current and fledgling treatment options.

INTRODUCTION

Primary immune thrombocytopenia (ITP), an immune disorder characterized by a reduction in the peripheral blood platelet count, arises from autoimmune hyperactivity. Typical symptoms include bleeding in the internal organs, mucous membrane, and skin, as well as fatigue and anxiety, but many individuals show few or no symptoms. In rare cases, fatal cerebral hemorrhage may appear. ITP has an annual incidence of 2–10/100 000 in adults, and is more frequent in women of reproductive age and the elderly over 60 years [1]. 

Box 1.

no caption available

ITP involves a combination of excessive destruction and low production of platelets [2]. The pathogenesis of ITP is summarized in Fig. 1[3^▪]. Due to the lack of a “gold standard” for diagnosis, ITP is often diagnosed in an exclusive manner [5^▪]: the presence of thrombocytopenia (platelet count <100 × 10⁹/l), the absence of anemia and leucopenia, and the exclusion of other causes of thrombocytopenia [2]. Patients manifesting a platelet count of ≥30 × 10⁹/l but no bleeding and a risk of ITP may be temporarily followed up with observation. To treat symptomatic patients, a platelet count should be maintained at >(20–30) × 10⁹/l [6]. Glucocorticoids are used to cope with ITP in the first line, but limited by an incidence of adverse effects over 20%. In the main second-line treatment, thrombopoietin receptor agonists (TPO-RAs), splenectomy, immunosuppressants can be adopted. The ITP treatment landscape flowchart is summarized in Fig. 2[1]. Rituximab therapy is preferred for patients who aim to avoid long-term medication and do not wish to undergo surgery [7]. However, rituximab cannot maintain its efficacy for a long term, leading to high risks of recurrence and adverse events. Besides, no indicators have been tested reliable to predict its efficacy, and special populations are not suitable for rituximab. In this study, we systematically analyzed the challenges in the treatment with rituximab, and explore the therapeutic potential of other possible candidates.

FIGURE 1.

The pathogenesis of ITP (original, created with BioRender.com). The onset of ITP is the result of both excessive platelet destruction and insufficient platelet production. This diagram explains it from the following five aspects: (1) The activity and number of cytotoxic T lymphocytes increase, which can directly mediate increased platelet clearance. The primary reasons are the imbalance of the Th1:Th2 ratio, the increased number of Th17 cells, and the deficiency of Treg cells, which are responsible for maintaining peripheral immune tolerance. (2) In the spleen, autoantibodies can bind to FcγR on macrophages, triggering phagocytosis. (3) Autoantibodies form complexes with platelets and simultaneously bind to complement C1, activating the complement cascade and promoting the formation of the membrane attack complex (MAC), leading to platelet lysis. (4) In addition to the spleen, aged platelets can also be cleared through the Ashwell–Morell receptor (AMR) in the liver. Autoantibodies can induce glycan modification of platelet surface glycoproteins (GPs), which can be recognized by AMR, leading to platelet clearance. This is referred to as desialylation of platelets. (5) Autoantibodies can bind to megakaryocytes, thereby causing inhibition of megakaryocyte production, maturation disorders, and reduced platelet release.

FIGURE 2.

ITP treatment landscape flowchart [adapted image source from Chinese Society of Hematology, Chinese Medical Association, Thrombosis and Hemostasis Group. Chinese Guidelines for the Diagnosis and Treatment of Adult Primary Immune Thrombocytopenia (2020 Edition). Chinese J Hematol 2020;41:617–23. doi:10.3760/cma.j.issn.0253-2727.2020.08.001].

RITUXIMAB FOR IMMUNE THROMBOCYTOPENIA

Rituximab, a chimeric monoclonal antibody, has a Fab segment specifically binding to target antigen CD20 and an Fc segment to a variety of receptors. It triggers antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) to cause the apoptosis and lysis of target cells, thereby reducing platelet antibody production and exerting immunosuppressive effects [8]. The mechanism of action of rituximab on B cell is summarized in Fig. 3 [9].

FIGURE 3.

The mechanism of action of rituximab on B cell (original, created with BioRender.com).

Currently, rituximab is a second-line option for ITP patients after the failure of glucocorticoid therapy [6]. In The Chinese Guidelines for the Diagnosis and Treatment of Primary Immune Thrombocytopenia in Adults (2020 Edition), the dosage of rituximab treatment is standardized as follows: 375 mg/m² i.v. once a week for four doses, with a short-term remission rate of 60–70% and an onset of action usually in 4–8 weeks [9]. In recent studies, a low-dose replacement therapy (100 mg or 100 mg/m² once a week for 4 weeks) has also been proposed. Compared to the standard regimen, this therapy is safer and cheaper, but shows no significant difference in efficacy, as shown by the complete remission rate (CRR) (RR 1.61 vs. 1.42, P = 0.45), overall remission rate (ORR) (RR 1.26 vs. 1.49, P = 0.28), partial remission rate (PRR) (RR 1.25 vs. 1.00, P = 0.11), and sustained remission rate (SRR) at month 12 (RR 2.00 vs. RR 1.64, P = 0.54). In addition, two other regimens, including the double standard-dose regimen (750 mg/m² once weekly for 4 weeks) and the “RA” regimen (1000 mg on days 1 and 15), can achieve the similar short-term efficacy to that of the standard dosage [11].

In a study evaluating the long-term safety and efficacy of rituximab, an initial response shows in 60.9% of the patients and the 5-year SRR in 30% [12], suggesting that the long-term efficacy of rituximab cannot be maintained. Another 5-year study shows an initial ORR of 57% in both adults and children, with the 5-year SRR of 21% and 26%, respectively [13]. Sang Baohua et al.[14] have explored the clinical efficacy of rituximab in children with ITP, showing that although the overall effective rate of rituximab is 70%, the recurrence rate is as high as 30% [12].

Long-term use rituximab also elicits resistance. A research [3^▪] details that the resistance may develop from the abnormal activities of B cells, T cells and plasma cells, represented by autoreactive anti-GpIIbIIIa antibody-secreting long-lived plasma cells (LLPCs) and the persistent high levels of B-cell activating factor (BAFF) in splenocytes. The abundance of CD4⁺ T cells, which can prolong plasma cell lifespan, increases in the peripheral blood of rituximab-resistant patients, so depletion of CD4⁺ T cells may hint at a way of addressing rituximab resistance. In addition, the remodeling of B cells and the increase of cytotoxic T lymphocyte (CTL) and follicular helper cell (TFH) are involved in the development of resistance to rituximab.

Predictors of response to rituximab

In a retrospective study using standard rituximab doses for ITP, younger women (age < 40 years) have a higher probability of experiencing R, CR, and long-term remission (LTR) [15]. Chapin et al. have shown a higher SRR in adolescent females with ITP <12 months than males. In another study based on the same treatment regimen, the LTR is higher in females of childbearing age with an ITP duration <24 months [16]. In addition, other studies suggest that younger women with a relatively shorter disease duration usually have better outcomes than other populations [12]. However, in several other studies, age and gender do not show a high predictive value [9].

The effect of antinuclear antibody (ANA) on the efficacy of rituximab remains unelucidated. Brah et al. have found a stronger response rate in ANA-positive ITP patients, compared to ANA-negative cases, but the difference is not statistically significant. ITP patients who test positive for ANA may have a better initial response to rituximab therapy, but lower SRRs than ANA-negative patients at 6, 12, and 24 months, suggesting that positive patients have a poor long-term outcome [17].

Recent studies have shown that ITP patients with anti-GPIIb/IIIa antibodies are more responsive to rituximab, but the RR does not show a significant association with the level of anti-GPIb/IX autoantibodies [18]. Given these controversial findings, antiplatelet autoantibodies are not currently recommended to be performed to predict the response to rituximab therapy [19]. Other predictors, such as bone marrow cell count and platelet count (tested at 14, 30, and 60 days after rituximab treatment), still need more studies to validate their predictive potential for long-term efficacy [20].

In summary, no clinical or experimental parameters are reliable to predict the response to rituximab, which hinders the precise use of rituximab.

Adverse effects of rituximab

The adverse effects of rituximab are infusion-related. Chills and fever are the most common during the first infusion, occasionally along with more severe manifestations, such as dyspnea, angioedema or hypotension. Serious adverse reactions include serum sickness, hepatitis B virus infection, or hypogammaglobulinemia [21]. Most of the serious infections occur within one year of rituximab use, and become rare years later, mainly in the elderly population on long-term steroids or immunosuppression.

Populations not suitable for rituximab

Antithyroid antibodies have been previously reported detectable in 18–39% of patients with ITP, and the positivity rate for antithyroid antibodies is higher in children with chronic ITP than in the general population. In a study using a low-dose rituximab for ITP in children, the children with positive antithyroid antibodies face a poor outcome, but the contributive mechanisms remain to be investigated [21]. In addition, ITP often causes maternal and secondarily neonatal thrombocytopenia in early and mid-pregnancy [23^▪]. Rituximab brings with teratogenicity during pregnancy [19]. The U.S. Food and Drug Administration labels rituximab as a class C drug, recommended only for pregnant patients with extremely serious symptoms [23^▪].

Rituximab is also associated with an increased risk of HBV reactivation and is, in principle, contraindicated in patients with active hepatitis B. However, in certain ITP patients comorbid with chronic HBV infection and deprived of other effective treatment options, rituximab may be administrated along with an antiviral therapy. Patients with prior HBV infection and whose symptoms are currently in remission should undergo HBV DNA load testing during months after rituximab treatment [1,5^▪,6].

ALTERNATIVES TO RITUXIMAB

Splenectomy

Splenectomy, an effective treatment for glucocorticoid-resistant ITP, can completely resolve platelet destruction and autoantibody production, with the highest durable response rate (50–70%) [24]. Moulis et al. are the first to directly comparing the outcomes of splenectomy and rituximab in ITP patients, showing that the RR, CRR, and maintenance rate of splenectomy are higher than those of rituximab [25]. Hammond et al. have found that the 2-year freedom from relapse (FFR) index was similar or even superior in patients receiving sequential splenectomy-rituximab or rituximab-splenectomy, compared to those treated with rituximab alone [26].

Drugs targeting to inhibit the FcγR signaling

In ITP, platelet destruction occurs as platelets are encapsulated by antibodies and then phagocytized. Therefore, the FcγR signaling pathway can be blocked to inhibit phagocytosis, thereby preventing platelet destruction. Spleen tyrosine kinase (Syk) and Bruton's tyrosine kinase (BTK) are key tyrosine kinases that play a role in B cell development, as well as macrophage FcγR-mediated phagocytosis [27]. Drugs targeting to inhibit the FcγR signaling may offer new therapeutic approaches for refractory ITP [8].

Spleen tyrosine kinase inhibitors

Fostamatinib, a selective Syk small molecule inhibitor, has been proposed as a novel agent for the treatment of ITP. This oral drug reduces antibody-mediated platelet destruction by inhibiting Syk to silence the B-cell receptor signaling. In two phase 3 trials, patients with ITP are randomly assigned to the Fostamatinib (a Syk inhibitor) group and the placebo group in a ratio of 2 : 1. About 18% of patients in the fostamatinib group demonstrate a stable response (6 visits every 2 weeks during the 14th to 24th week of treatment, platelet count ≥50 × 10⁹/l in at least 4 times of testing, and no remedial therapy), compared with the only 2% in the placebo group. OR (platelet counts ≥50 × 10⁹/l at least once during the first 12 weeks of treatment) is achieved in 43% of patients in the treatment group, but only 14% in the placebo group. Most adverse events reported in the study, mainly diarrhea and hypertension, are mild to moderate, and can be relieved spontaneously or with medication [27].

Bruton's tyrosine kinase inhibitors

Like Syk, BTK is involved in FcγR signaling and considered as another fighter against ITP [2]. Rilzabrutinib is an oral, reversible, covalent, small-molecule BTK inhibitor. As high as 64% of patients in the rilzabrutinib treatment group can experience a response within the first 12 weeks of treatment (platelet count ≥50 × 10⁹/l or between 30 and 50 × 10⁹/l), compared with only 32% in the placebo group. The 23% DRR in the treatment group is sufficient to demonstrate the superiority of rilzabrutinib, considering that no one shows a durable response (platelet count ≥50 × 10⁹/l for at least two-thirds of the 12–24 weeks of treatment in the absence of emergent treatment) in the placebo group. Adverse effects are predominantly mild to moderate in the rilzabrutinib group, indicating a favorable safety profile [28^▪].

Neonatal Fc receptor inhibitors

The neonatal Fc receptor (FcRn), so named because it was originally discovered in the neonatal gut of rodents, plays a key role in regulating the half-life of immunoglobulin G (IgG) and albumin [29]. It binds to IgG to avoids its degradation, and releases it from endothelial endosomes, till recirculates it to the cell surface. FcRn recycles normal IgG as well as pathological IgG in the same way [30]. Therefore, FcRn inhibitors can block the FcRn–IgG binding to promote the degradation and shorten the half-life of IgG [31].

Rozanolixizumab, a human monoclonal anti-FcRn antibody that can be infused subcutaneously, significantly reduces IgG levels in patients’ serum. In a trial including 66 adult ITP patients, 66.7% and 54.5% of patients display platelet counts ≥50 × 10⁹/l at least once after receiving a single subcutaneous infusion of 15 mg/kg and 20 mg/kg rozanolixizumab, respectively, and the adverse events (mainly headache) are mild, suggesting the good safety and tolerability profile of this inhibitor [31].

In a phase 2 study including 38 patients with ITP, the risk of infection does not increase significantly and the adverse events are only mild to moderate in the efgartigimod-treated group, compared to the placebo group. A platelet count ≥50 × 10⁹/l in just one testing is achieved in a higher number of individuals receiving high doses of placebo (50% vs. 14%), but a sustainable count can only rely on efgartigimod. About 38% of patients intravenously given either 5 or 10 mg/kg efgartigimod demonstrate a durable response (platelet count ≥50 × 10⁹/l for at least 10 consecutive days), whereas none in the placebo group [32].

Thrombopoietin receptor agonist

To avoid the adverse effects of immunosuppressive drugs, the first generation of thrombopoietin (TPO)-mimetic drugs was developed in the 1990s, including recombinant human thrombopoietin (rhTPO) and polyethylene glycol-modified recombinant human megakaryocyte growth and development factor (PEG-rhMG-DF). The patients’ platelet counts increase upon the stimulation by the TPO receptor. Despite the promising efficacy of these mimics against ITP, some patients can develop antibodies against the drug, and because of the similarity of the TPO mimics to natural endogenous human TPO, the antidrug antibodies may cross-react with the natural TPO, ultimately leading to severe thrombocytopenia. As a result, the first-generation TPO mimics has been rejected in clinical use [6].

Second-generation TPO-receptor agonist (RAs), including romiplostim and eltrombopag, have transformed the treatment of ITP [33,34^▪]. These TPO-RAs have no similarities to natural thrombopoietin, so any antibodies directed against the drug do not cross-react with the patient's native thrombopoietin, thereby reducing the incidence of serious adverse reactions. As a second-line therapy, TPO-RAs have shown superior efficacy and safety to rituximab or splenectomy [7].

Both romiplostim, a peptide that directly competes for binding to the TPO binding site, and eltrombopag, a small molecule that binds at the transmembrane site, can bind to cause a conformational change in the TPO receptor, consequently boosting megakaryocyte proliferation and platelet production. A meta-analysis shows a significant therapeutic effect of TPO-RAs, as they can increase the platelet response rate by threefold and DRR nearly 8-fold compared to placebo [35]. Despite this therapeutic effect, the drug should be timely switched in the case of obvious platelet count fluctuations or side effects. Patients who do not show a response to one TPO-RA may show a response to an alternative. In a trial of 46 TPO-RA drug-switching patients, 46% of romiplostim nonresponders show responses after switching to eltrombopag. Similarly, 80% of eltrombopag nonresponders turn to respond to romiplostim [36]. In addition, approximately 10–30% of patients can maintain a response after discontinuing TPO-RAs [37].

Plasma cell/B-cell targeted inhibitors

Proteasome inhibitors (bortezomib) induce the apoptosis of plasma cells and memory B cells by inhibiting the ubiquitin-proteasome protein waterway channel, thus reducing the secretion of antibodies [3^▪]. Similarly, monoclonal antibodies against plasma cell membrane glycoproteins, such as anti-CD38 antibodies (daratumumab and mezagitamab) or anti-CD19 antibodies (inebilizumab and obexelimab), have an ability to deplete cloned plasma cells.

Mezagitamab, a human IgG1λ monoclonal antibody, binds to CD38 to deform the enzyme and inhibit its activity, thereby inducing cellular apoptosis. Besides, it can lyse cells via ADCC and CDC to decrease the abundances of NK cells as well as B- and T-lymphocyte subpopulations [38]. Due to this mechanism, mezagitamab is associated with such complications as hypogammaglobulinemia and even serious infections. Nevertheless, its safety and efficacy have been confirmed in a phase 1 trial [39].

Obexelimab is an anti-CD19 monoclonal antibody containing an Fc fraction with a high affinity to FcγRIIB expressed by B-cells, while FcγRIIB is responsible for inhibiting B-cells and reducing the production of pathogenic antibodies. Thus, the Fc portion of this drug binds to FcγRIIB expressed on B cells, and the Fab portion recognizes CD19 expressed by B cells, thereby reducing pathogenic antibodies. Since ITP is an antibody-mediated immune disease, the use of obexelimab deserves full caution [40].

As mentioned earlier, rituximab resistance arises in part due to the persistence of long-lived plasma cells in a BAFF-rich environment. In addition, anti-BAFF antibodies in combination with anti-CD20 treatment can lead to a massive depletion of long-lived plasma cells in a mouse model. Therefore, BAFF as an important target may be inhibited for mitigating rituximab resistance. Belimumab, as a human IgG1λ monoclonal antibody against BAFF, causes a reversible reduction in B-lymphocyte production. Combining rituximab with five infusions of belimumab achieves a 1-year ORR of 80% and a CRR of 66.7%, both of which are higher than those in the use of rituximab alone (1-year ORR of 40–50% and CRR of 30%) [41].

T-cell targeted inhibitors

Some immunosuppressants, such as decitabine and cyclosporin, can be used to reduce platelet destruction by regulating the number and function of T-lymphocytes.

Previous studies have shown that low-dose decitabine promotes platelet production, restores Treg function, and inhibits CTL cytotoxicity onto platelets. In a study evaluating the efficacy and safety of low-dose decitabine in patients with ITP, the intravenous infusion of 3.5 mg/m² decitabine (three consecutive days per cycle, three cycles with a 4-week interval) achieves a CRR of 17.78%, a PRR of 33.33%, and SRRs of 44.4%, 31.1% and 20% at 6, 12, and 18 months, respectively. Among the responders, immune thrombocytopenia-related symptoms, fatigue, mental health, and fear are all alleviated, the overall quality of life is significantly improved, and no serious adverse events are recorded [42].

Cyclosporin is a widely used potent inhibitor on both humoral and cellular immunity. It inhibits TFH and CTL activity to prevent platelet destruction. A meta-analysis shows that the combination regimens of cyclosporin can increase the ORR and CRR, decrease the relapse rate, but do not induce more adverse drug reactions [43^▪].

Complement inhibitors

Complements play an important role in the pathogenesis of ITP; however, the efficacy of treatments targeting the complement pathway is unclear [6]. Complement inhibitors are promising therapeutic agents for ITP.

In a phase I trial of 12 patients with chronic ITP who have received at least two failed treatment regimens (including eight patients responding inadequately to rituximab), sutimlimab, a humanized C1s IgG4 monoclonal antibody, exhibits mild adverse effects (mainly fatigue, no deaths or thromboembolic events), and no patients discontinue treatment due to drug-related adverse events. A durable platelet count response is achieved in five patients (a platelet count maintained at ≥50 × 10⁹/l in ≥50% of follow-up recordings), and CR in 4 patients (two consecutive platelet counts ≥100 × 10⁹/l with ≥7 days between two recordings) [4,10,22,34^▪].

CONCLUSION

In the treatment of ITP, the high remission rate of rituximab cannot persist long, unavoidably followed by resistance or relapse. Rituximab is associated with some adverse effects, and there are no reliable predictors for treatment response. In addition, its use is limited in some special populations. Therefore, other candidates are still needed to overcome the shortcomings of rituximab. Splenectomy after rituximab is more effective than rituximab alone, and novel therapies targeting the FcγR, FcRn, thrombopoietin receptor, B-cells or plasma cells, T-cells, and the complement pathways have shown an encouraging efficacy in patients demonstrating poor outcomes after the treatment with rituximab.

Acknowledgements

This work was supported by Program of Taizhou School of Clinical Medicine of Nanjing Medical University (Nos. TZKY20220310 and TZKY20240201).

We express our deepest gratitude to Professor Jianfeng Zhu for his invaluable guidance. Furthermore, we extend our heartfelt thanks to Teacher Yongke Cao for the guidance on the language expression of the thesis and continuous support throughout this research. We thank BioRender (BioRender.com) for providing the graphical tools used in this study.

Financial support and sponsorship

This work was supported by Program of Taizhou School of Clinical Medicine of Nanjing Medical University (Nos. TZKY20220310 and TZKY20240201).

Conflicts of interest

There are no conflicts of interest.

REFERENCES AND RECOMMENDED READING

Papers of particular interest, published within the annual period of review, have been highlighted as:

• ▪ of special interest

• ▪▪ of outstanding interest