Successful treatment of refractory secondary immune thrombocytopenia (antiphospholipid antibody syndrome-associated) with the combination of rituximab and romiplostim at the cost of severe bone pain: A case report and review of literature

Introduction

Immune thrombocytopenia (ITP) is an autoimmune disease characterized by isolated thrombocytopenia of <100 × 10⁹/L. The incidence of ITP is 2–4/100,000 adults of which secondary ITP (associated with infections, malignancies, autoimmune diseases) constitutes ∼20%. ¹ The underlying pathophysiology is not entirely understood. There are several proposed mechanisms. First and foremost, peripheral platelet destruction due to antibodies against platelet membrane components and platelet-reactive T-cells which lead to decreased platelet life span and increased turnover. Second, megakaryocyte destruction in the bone marrow due to antibodies against megakaryocytes and autocytotoxic T-cell activation in the bone marrow resulted in impaired platelet production. Other suggested mechanisms include megakaryocyte viral infections and thrombopoietin deficiency.^2,3

Treatment is mainly considered for patients with platelet counts <20–30 × 10⁹/L or bleeding symptoms. Corticosteroids are considered as the standard of care first-line treatment options (±intravenous immunoglobulin (IVIG) or Anti-D as an alternative).^2,4,5 Patients with ITP recurrent or refractory to corticosteroids are evaluated for the second-line agents or splenectomy based on patients’ clinical status and agents’ adverse effect profiles. Suggested second-line and subsequent agents include thrombopoietin receptor agonists (TPO-RAs: romiplostim and eltrombopag), rituximab, immunosuppressive therapy (azathioprine, cyclophosphamide, cyclosporin A, and mycophenolate mofetil; alone or combined), and others (danazol, dapsone, vinca alkaloids, recombinant interferon alfa2b, and fostamatinib).^2,4,6

To our knowledge, there are a total number of four case reports in the literature, describing a total of 10 patients, where the combination of romiplostim and rituximab was used. This combination significantly increased the platelet counts in those with ITP refractory to standard treatments. It induced long-term remission in 9 out of 10 patients without the need for romiplostim maintenance therapy.^7–10 However, this combination’s safety and adverse effects have not been widely discussed. We describe a 40-year-old male patient successfully treated with the combination of romiplostim and rituximab who suffered from grade 3 extremity and joint pain and refused to continue treatment.

Case report

On 11 April 2014, 40-year-old male patient who was diagnosed with ITP in 2010 presented with right popliteal deep venous thrombosis. Complete blood count showed no abnormalities. He was placed on warfarin with heparin bridge therapy. Antiphospholipid IgG > 100 U, Anti B2-glycoprotein IgG > 100 U, IgM > 35 U; lupus-anticoagulant was negative. Factor V Leiden mutation, prothrombin gene mutation was negative, MTHFR mutation was heterozygous positive. Protein C and S were low (while the patient was on warfarin). He showed no clinical features of systemic lupus erythematosus. Anti-Nuclear Antibody, Anti-dsDNA, Anti-Sm were negative. C3, C4 levels were normal. More than 12 weeks after the initial testing, anti-phospholipid antibody syndrome (APLAS) workup was repeated. Anti-phospholipid IgG > 100 U, Anti B2-glycoprotein IgG > 100 U, IgM: 28.5 U. The patient was diagnosed with APLAS.

Management and outcome

On 23 April 2018, he presented with right lower extremity swelling for the last two to three days with generalized purpuric rash and oral mucosal bleeding. He was on warfarin with international normalized ratio (INR): 1.7 and his platelets (Plt) had dropped down to 14 × 10⁹/L (previously 150 × 10⁹/L). He was admitted and administered methylprednisolone (1 mg/kg) with an improvement of Plt: 45 × 10⁹/L on day 4 (D4). He was discharged on prednisone and low-molecule-weight heparin Q12 on D4. On D8, platelets dropped to 7 × 10⁹/L. On D9, he was readmitted and was administered IVIG (400 mg/kg/day for five days), prednisolone was continued.

Due to partial response (Plt: 23 × 10⁹/L on D16) after the fifth dose of IVIG, he was started on romiplostim (2 µg/kg) weekly on D17. On D29 Plt: 14 × 10⁹/L, the third dose of romiplostim was increased (4 µg/kg). On D60, due to lack of response to romiplostim (Plt: 26 × 10⁹/L after seventh dose), bone marrow biopsy was performed: “Cellular bone marrow biopsy at %50 with trilineage hematopoiesis including megakaryocyte hyperplasia with a complete morphological spectrum likely related to romiplostim.” Flow cytometry, cytogenetics, and Myelodysplastic Syndrome Fluorescence in-situ hybridization panel results were within normal limits, Hepatitis B virus and Hepatitis C virus serology, and PNH testing were negative.

On D67, romiplostim (eighth) dose was increased (to 6 µg/kg) and rituximab (375 mg/m²) QW was started on D68. Rituximab second dose was administered on D72. On D74, the patient presented to the emergency department with Grade III leg pain and an increase in Plt: 162 × 10⁹/L (from 43 × 10⁹/L on D71) and romiplostim was held. The third dose was given on D82 and the fourth dose of rituximab was administered on D86. On D92, Plt count decreased to 28 × 10⁹/L. He was reinitiated on Rituximab (fifth dose) with Romiplostim (6 µg/kg) (D92 and D93, respectively). Five days later, on D98 he presented to the emergency room with grade III back pain. On D100 rituximab sixth dose and romiplostim were administered. The next day he presented to the ER with Grade III lower extremity, sacral, lumbar and chest pain. After the differential diagnosis, the pain did not improve with intravenous hydromorphone, intravenous corticosteroids, oral loratidine (given for suspected growth-factor-related bone pain). ITP treatment was stopped and the patient was seen in monthly follow-ups for approximately one year up to this date.

Discussion

ITP is a disease characterized by increased platelet destruction and decreased platelet production. The main second-line agents are rituximab and TPO-RAs. Rituximab has an overall response rate (ORR) of 60% with a median time to response (TTR) of one to eight weeks. It is preferred commonly because of the potential for long-term remission. Twenty-one percent of adults that responded at one year had durable responses at five years. ¹¹ This long-term durable response, however, may be due to the natural course of the disease. TPO-RAs had an 80% ORR with a median TTR of 10–14 days. These agents have high rates of sustained response to continuous treatment. ¹¹ They are preferred when safe platelet levels and a lower risk of bleeding is needed more promptly, especially for a splenectomy, other major surgeries, and in cases of major bleeding.

Several reports suggest that refractory cases could benefit from the combination of rituximab and romiplostim.^7–10 It is speculated that the two different underlying pathophysiologies—immune-mediated platelet destruction and insufficient production—could be targeted by agents with mainly two different mechanisms of action in ITP. The downregulation of autoimmunity by rituximab together with the enhanced platelet production by means of romiplostim could, therefore, demonstrate an increased efficacy. Furthermore, TPO-RAs stimulate an expeditious increase in platelets which is reliant on the continuity of the therapy, whereas rituximab promotes a slower response with a potential of long-term remission. Thus, the combination may create a complementary effect, which may result in continually desired platelet levels. Moreover, it is advocated that TPO-RAs also regulate the immune system via Tregs and Bregs. ¹² This may also complement the additive effect.

In this patient’s case, platelet count responded aptly every time romiplostim dose was increased. The fact that the platelet count decreased drastically when romiplostim was paused proves that romiplostim has an unequivocal influence on this patient’s platelet count. On the other hand, 24 days after the rituximab administration (TTR = one to eight weeks), platelet count started a steady ascent which, on the contrary to previous levels, was not impeded after the patient’s wish to stop romiplostim. This durable platelet response may be supporting evidence of rituximab’s late-onset efficacy (Figure 1).

OPEN IN VIEWER

Figure 1.

Treatments given and platelet count over time. The increase in platelet counts become durable after rituximab administrations. Larger orange dots: Romiplostim administration, black arrows: Rituximab (375 mg/m²) administration. The diagram depicts platelet counts (blue line).

Maloney et al. reported that 10% of the patients (n:2/20) treated for non-Hodgkin lymphoma with weekly rituximab doses up to 375 mg/m² experienced grade 3 arthralgias. ¹³ Kuter et al. reported that a total of 23% of patients experienced serious adverse events from romiplostim (grades 3–4), where extremity pain, back pain, and arthralgias constitute a total of 3%. ¹⁴ Our patient has experienced serious leg, back, and hip pains every time he was administered the combination of rituximab and romiplostim whether simultaneously or one day apart, on three separate occasions.

Pohlen et al. reported administering this combination to two patients. The first patient received a dose of rituximab one day after romiplostim on a single occasion. The second patient received four doses of rituximab, weekly, every two days before romiplostim. No painful episodes were reported for any of the patients. ⁷ Veneri et al. reported using the combination on a single patient as four doses of rituximab weekly with romiplostim. There were not any painful episodes that were noted. ⁸ Contis et al. had a larger patient population where four patients were treated simultaneously with rituximab and romiplostim. One of the patients stopped taking the combination after the second dose after reaching high platelet counts (>500 × 10⁹/L), the rest of the patients received four doses of rituximab simultaneously with romiplostim weekly. Three patients (75%) had durable responses. Only one patient had arthralgia which led to dose reduction from 4 μg/kg/week to 2 μg/kg/week. ⁹ Merono et al. used the combination on a single patient where the four doses of rituximab were administered two days after romiplostim infusion. He experienced lumbar pain which responded to analgesic treatment. ¹⁰ Including the patient we described, ORR for patients treated with the combination of rituximab and romiplostim was 100% (n:10/10) where sustained response without the need to continue romiplostim was seen in 90% (n:9/10). On the other hand, 30% (n:3/10) suffered from grades 3–4 arthralgias and/or bone pains, requiring analgesics. When calculated for the combination, the Naranjo score is 7 which infers a high probability. However, when calculated for romiplostim which already has conclusive reports on limb pain, the Naranjo score is −2 which means a reaction caused by romiplostim was unlikely.

We acknowledge that this patient received rituximab doses four days apart on two occasions (on D68-72 and D82-86) which is shorter than the proposed one-week intervals. Another deviation from the recommended regimens is that this patient received six doses of rituximab instead of the suggested four doses. This increase in serum steady-state drug concentration might have an effect on the adverse event frequency. However, Maloney et al. reported that the first rituximab administration causes the majority of the adverse events which is speculated to be a reaction to the destruction of circulating B-cell mass. From doses two to four, there was a scarcity of adverse events. ¹³ Furthermore, according to Keating et al.’s study, doses as high as 2.250 mg/m² result in neither any unusual adverse events (predominantly grade 2 infusion-related reactions) nor any greater than grade two adverse events. ¹⁵ These data contradict any effect of medication buildup on adverse effect frequency or profile.

Other studies that have studied the efficacy of rituximab combination treatment with either eltrombopag or recombinant human trombopoietin (rhTPO) also failed to show a clear association with bone/joint pain. Gómez-Almaguer et al. used low-dose (100 mg/m²) rituximab, eltrombopag and dexamethasone combination on 13 patients which only resulted with one patient (%7) having mild myalgia. ¹⁶ Li et al. also used low-dose rituximab (100 mg/m²) combined with rhTPO in 14 patients with no myalgia/arthralgia encountered. ¹⁷ Another randomized open-label study which compares standard dose rituximab (n =38) to low-dose rituximab plus rhTPO combination (n = 77) also failed to show significant difference in bone/joint pain adverse event incidences (5.3% vs. 3.9% respectively).¹⁸ It must be underlined that all three groups used low-dose rituximab, which may or may not have an effect on the medication synergism in terms of arthralgia adverse events.

Conclusions

We reason that the mitigation of platelet and megakaryocyte destruction by rituximab along with the expansion of platelet production by romiplostim may act synergistically in select cases where a faster and more sustainable improvement is needed. However, we hypothesize that this combined effect may also cause a substantial hyper-proliferation in the bone marrow, causing serious grades 3–4 bone pains. Additionally, lower dose (100 mg/m²) rituximab combinations may prove to be more secure, which should be considered in future studies. Since data gathered from a small number of patients that have been reported is limited; randomized controlled trials on a larger cohort to validate the safety and efficacy of this combination is needed.