Abstract
The management of pregnant women with immune thrombocytopenia who fail to respond to corticosteroids and intravenous immunoglobulin is an intractable clinical challenge because of the limited availability of evidence-based information. Recombinant human thrombopoietin (rhTPO) is recommended for refractory immune thrombocytopenia (ITP). To date, however, few studies have investigated rhTPO treatment during pregnancy. We retrospectively reviewed four cases who were diagnosed with ITP and treated with rhTPO during pregnancy in our center from January 2015 to June 2020. Of the four cases, two (50%) responded to rhTPO treatment. No adverse events were noted in the newborns. Our findings indicate that rhTPO treatment is safe for patients with refractory gestational ITP, and that subcutaneous injection is a convenient delivery method that does not lead to adverse events. Thus, rhTPO may be a viable alternative treatment option for patients with refractory gestational ITP who do not respond to first-line therapies.
Keywords
Introduction
Thrombocytopenia is a common clinical disease that is characterized by platelet counts of less than 150 × 109/L. Approximately 5% to 10% of pregnant individuals exhibit reduced platelet counts during pregnancy. 1 The causes of thrombocytopenia vary across the trimesters of pregnancy (Figure 1). In the first trimester, over 75% of thrombocytopenia is caused by immune thrombocytopenia (ITP) or hereditary thrombocytopenia (HT). In the second and third trimesters, platelet counts tend to receive more attention. During the second trimester, the most common reasons for platelet counts >100 × 109/L are gestational thrombocytopenia, ITP, PEC, and other diseases, whereas ITP is regarded as the major cause of platelet counts < 100 × 109/L. In the third trimester, the main causes of platelet counts both >100 × 109/L and <100 × 109/L are gestational thrombocytopenia and PEC. 1 Of these, ITP accounts for 3% to 5% of all cases of gestational thrombocytopenia, and is the most common cause of platelet counts below 50 × 109/L during the first and second trimesters.1–8 Pregnant individuals with severe thrombocytopenia (platelet counts <20 × 109/L) are at high risk of essential hemorrhage, postpartum hemorrhage, and placental abruption. 9 Among patients with gestational ITP (P-ITP), approximately one-third are diagnosed during pregnancy and two-thirds are diagnosed before pregnancy. 10 The children of mothers with ITP may also exhibit thrombocytopenia because maternal anti-platelet immunoglobulin G antibodies can pass across the placenta, resulting in platelet damage in the neonate.3,5 However, this is considered to be a transient condition and does not lead to hemorrhage in the fetus.

Prevalence of causes of thrombocytopenia based on trimester of presentation and platelet count. The size of each circle represents the relative frequency of all causes of thrombocytopenia during each of the three trimesters of pregnancy. All etiologies and all platelet counts are considered together in the first trimester when thrombocytopenia is uncommon. Distribution of etiologies during the second and third trimesters is subdivided by platelet count. All results are estimates based on personal experience and review of the literature. “Other” indicates miscellaneous disorders, including infection, disseminated intravascular coagulation, type llB von Willebrand disease, immune and nonimmune drug-induced thrombocytopenia, paroxysmal nocturnal hemoglobinuria, and bone marrow failure syndromes (aplastic anemia, myelodysplasia, myeloproliferative disorders, leukemia/lymphoma, and marrow infiltrative disorders). Figure and legend reproduced from 1 with kind permission from the American Society of Hematology.
Regardless of pregnancy status, oral glucocorticoids and intravenous immunoglobulin (IVIg) are the primary therapeutic options for ITP. For patients with non-gestational ITP, the response rates to glucocorticoids and IVIg are 80% 11 and 75%, 12 respectively. However, Sun et al. 13 reported response rates to glucocorticoids and IVIg of 39% and 38% in patients with P-ITP, indicating markedly reduced response rates in this population. Notably, once first-line treatments fail, few second-line treatments are available for individuals with P-ITP.
Thrombopoietin (TPO) exerts its biological effects by binding its specific receptor (myeloproliferative leukemia protein [Mpl]), thus regulating the proliferation, differentiation, maturation, and division of megakaryocytes into functional platelets. Recombinant human TPO (rhTPO), a full-length glycosylated TPO that is expressed in Chinese hamster ovary cells, is a recombinant ligand of c-Mpl. 14 In China, rhTPO has been approved by the State Food and Drug Administration for refractory ITP with hormone ineffectiveness. A previous multicenter randomized trial found that rhTPO in patients with refractory ITP with hormone ineffectiveness had a high response rate (of 60.3%), good tolerance, and a significantly reduced risk of bleeding events. 15 Although the use of rhTPO for the treatment of P-ITP has not yet been extensively studied, existing evidence suggests that rhTPO is a safe and effective treatment for pregnant women with thrombocytopenia. In the present study, we therefore investigated four cases of P-ITP who were treated with rhTPO in our center from January 2015 to June 2020, and conducted a related literature review. The reporting of this study conforms to CARE guidelines (for CAse REports). 16
Case presentations
Patient 1
A 33-year-old woman diagnosed with ITP underwent her first pregnancy. Her baseline platelet count was around 50 × 109/L. Because of her ITP history, the patient underwent routine blood examinations every 2 to 3 weeks during pregnancy in our department (Department of Hematology, Affiliated Hangzhou First People’s Hospital, Zhejiang Chinese Medical University). At 15 weeks of pregnancy, her platelet count dropped to 30 × 109/L, and she was administered methylprednisolone (1 mg/kg/day). After 7 days, her platelet counts had not significantly increased. Following the administration of 400 mg/kg/day of IVIg for 5 days, her platelet count increased to 44 × 109/L. Less than 1 week later, at 17 weeks of gestation, the patient developed severe bleeding of the gums and skin, and her platelet count dropped to 7 × 109/L. Platelet infusion was performed for 2 to 3 weeks to maintain platelet levels. In week 24 of pregnancy, her platelet count decreased to 8 × 109/L. Considering that the patient had failed to respond to corticosteroids, it was suggested that she undergo 300 U/kg/day of rhTPO for 7 days; she accepted this advice. After 7 days of treatment, her platelet count had increased to 14 × 109/L. Her platelet counts were then maintained until 32 weeks of pregnancy by platelet transfusion every 1 to 2 weeks. She delivered via cesarean section at 32 weeks of gestation, and received a platelet transfusion before parturition. There was no hemorrhage in the placenta. She lost 500 mL of blood and did not need blood or platelet transfusion during the delivery. There was no blood loss after delivery. The weight of the preterm infant was 1840 g, and 1-minute/5-minute Apgar scores were 9/9. At 7 days postpartum, the condition of the newborn was good. The characteristics of all patients are shown in Table 1. The platelet change curves and treatment processes during pregnancy are shown in Figure 2.
Summary of information from the four patients.
C, corticosteroid; CR, complete response (platelet count of at least 100 × 109/L and the absence of bleeding); d, day; IVIg, intravenous immunoglobulin; ITP, immune thrombocytopenia; NR, no response (platelet count below 30 × 109/L); PT, platelet transfusion; R, response (platelet count more than 30 × 109/L and increase of at least twice that of the baseline).

Changing curves of platelet counts from Patients 1 (a), 2 (b), 3 (c), and 4 (d)
represents corticosteroids;
represents intravenous immunoglobulin;
represents platelet transfusion;
represents the use of rhTPO.
Patient 2
A 27-year-old woman with ITP and a baseline platelet count of 60 × 109/L experienced her first pregnancy. She was monitored in the outpatient section of the Department of Obstetrics and Hematology every 2 to 3 weeks. At 23 weeks of pregnancy, her platelet count decreased to 12 × 109/L. She was therefore administered 1 mg/kg/day of methylprednisolone. At 24 and 25 weeks, this therapy was repeated twice because of reduced platelet numbers. No effects of these corticosteroid treatments were detected; her platelet counts did not increase significantly. At 27 weeks of gestation, her platelet counts reduced to 9 × 109/L. After fully informing the patient of the advantages and disadvantages of rhTPO in pregnancy, the patient agreed to receive 300 U/kg/day rhTPO for 7 days. After 7 days, at drug withdrawal, her platelet count had increased to 32 × 109/L. Her platelet counts remained at 20 to 30 × 109/L until week 32 of gestation. At 32 weeks, large ecchymosis appeared in both lower limbs. After dexamethasone administration to promote fetal lung maturation, her platelet counts were monitored weekly, and delivery was completed after platelet transfusion at 37 weeks. A cesarean delivery was performed without any signs of placental hemorrhage. The total blood loss during the delivery was recorded as 1000 mL, and no blood transfusion was required. There were no further instances of post-delivery blood loss. The weight of the infant was 3090 g, and 1-minute/5-minute Apgar scores were 10/10. At 7 days postpartum, the condition of the newborn was good.
Patient 3
A 27-year-old woman with ITP had her first pregnancy. Her baseline platelet count was 40 × 109/L. Routine blood examinations were performed every 2 to 3 weeks. Until week 24 of pregnancy, her platelet counts fluctuated around 30 × 109/L with no serious bleeding events. At 27 weeks, her platelet count dropped sharply to 5 × 109/L; after treatment with 1 mg/kg/day of methylprednisolone combined with 400 mg/kg/day of IVIg for 5 days, her platelet count recovered. At 28 weeks, her platelet count reduced to 9 × 109/L, and she underwent the same treatment. However, at 29 weeks, her platelet count reduced to 8 × 109/L. Given the patient’s poor responses to corticosteroid therapy and her anxiety, 300 U/kg/day of rhTPO for 7 days was suggested, and she agreed to undergo this treatment. At 30 weeks, her platelet count was 10 × 109/L. Following delivery at 37 weeks, the patient was repeatedly treated with 400 mg/kg/day of IVIg for 5 days to consolidate her platelet count. Platelet transfusion was administered to the patient prior to delivery via cesarean section, and no placental bleeding was observed. During the delivery, the recorded blood loss was 600 mL; no additional blood or platelet transfusion was required. There was no further blood loss after delivery. The weight of the infant was 3530 g, and 1-minute/5-minute Apgar scores were 10/10. At 7 days postpartum, the condition of the newborn was good.
Patient 4
A 28-year-old patient with ITP and a baseline platelet count of 40 × 109/L experienced her first pregnancy. She had undergone splenectomy in 2014 and recovered well. At 12 weeks of pregnancy, she exhibited obvious gum and skin bleeding, and her platelet count was 2 × 109/L. Blood transfusion and administration of 400 mg/kg/day of IVIg were performed for 5 days, and her platelet count recovered. At 20 weeks, her platelet count again dropped to 8 × 109/L. She was administered 400 mg/kg/day of IVIg combined with 1 mg/kg/day of methylprednisolone; however, her platelet count did not increase significantly. At 21 weeks, she underwent platelet infusion and 400 mg/kg/day of IVIg for 5 days. At 22 and 28 weeks, she was again treated with 400 mg/kg/day of IVIg for 5 days because of low platelet counts. At 29 weeks, her platelet count was 12 × 109/L. However, the patient expressed a desire for more sustained and stable platelet levels to ensure a safe delivery. Given the failure of her responses to methylprednisolone and IVIg, 300 U/kg/day of rhTPO for 14 days was recommended, and she agreed to receive this treatment. At 31 weeks, her platelet count was 56 × 109/L; it remained at 50 × 109/L until 37 weeks. She did not receive blood or platelet transfusion before delivery. The delivery method was cesarean, and no bleeding occurred in the placenta. The patient lost 500 mL of blood during delivery and 300 mL of blood after delivery. No blood or platelet transfusion was performed. The weight of the infant was 3640 g, and 1-minute/5-minute Apgar scores were 10/10. At 7 days postpartum, the condition of the newborn was good.
Discussion
Clinically, ITP is regarded as an acquired autoimmune disease that is characterized by a transient or continuous decrease in platelet counts. 17 In the general population and in pregnant individuals, the incidences of ITP are 5 to 10 and 10 to 100 per 100,000, respectively.18,19 A patient’s immune intolerance against autoantigens mediates P-ITP. This disease is similar to adult ITP, including an antibody-induced increase in platelet destruction, a decrease in platelet production by megakaryocytes, and an immune mechanism that involves T cells. 7
A national cohort study in the UK 20 revealed an incidence of severe P-ITP of 0.83/10,000. The incidences of postpartum hemorrhage (51%) and severe postpartum hemorrhage (21%) in pregnant individuals with ITP are higher than those in pregnant individuals in general (5%–10%), but this bleeding is generally controllable, without adverse events. The mortality rate of mothers with ITP is <1%, the neonatal intracranial hemorrhage rate is <1.5%, and only 10% of neonates exhibit platelet counts below 50 × 109/L. 18 Given that blood is hypercoagulable during pregnancy, pregnancy is associated with a higher tolerance to ITP, and the pregnancy process is relatively safe.
The clinical features of P-ITP are comparable to those of non-gestational ITP. They include thrombocytopenia and skin and mucosal bleeding, and the severity of bleeding is similar to that of thrombocytopenia. The management of bleeding risk has traditionally been the main focus of ITP treatment. Treatment of P-ITP also aims to prevent bleeding, and P-ITP patients are generally treated when platelet counts are < 20 to 30 × 109/L or when severe bleeding occurs.21,22 Vaginal delivery is allowed in P-ITP pregnancies with platelet counts > 50 × 109/L. 23
Oral corticosteroids and IVIg are the first-line treatment options for P-ITP, the same as for adults with non-gestational ITP. A recent study 24 highlighted a combination of prednisone and IVIg as a potential combination therapy for pregnant individuals with ITP. Second-line treatment options for adults with non-gestational ITP include TPO receptor agonists (TPO-RAs), rituximab, and splenectomy. 25 Because of their definite teratogenic effects, danazol, mycophenolate mofetil, vincristine, and cyclophosphamide are not used in the gestational population. Moreover, rituximab can result in B cell dysfunction in neonates within the first year of birth, and is therefore not recommended for the gestational population. 26 TPO-RAs, including romiplostim and eltrombopag, are recommended for refractory ITP. However, there is limited evidence of the use of TPO-RAs during pregnancy, and some reports indicate that the use of TPO-RAs should be prohibited during pregnancy. 7 Conversely, some scholars consider that TPO-RA use during pregnancy can increase the response rates of ineffective platelet therapy and reduce the risk of bleeding. 27
For refractory ITP, rhTPO is recommended.28–30 However, few studies have investigated rhTPO treatment during pregnancy. After multiple drug treatments, platelet counts do not increase in some pregnant individuals, leading to increased bleeding risk. Although the data on rhTPO treatment during pregnancy are mostly limited to individual reports, rhTPO has been shown to have considerable safety and efficacy among these limited cases.31–35
A previous in vivo study 36 revealed that rhTPO effectively increases platelet counts in P-ITP mouse models, with negligible side effects on pregnancy. Similarly, a recent study 37 reported that 23 of 31 (74.2%) pregnant individuals exhibited positive responses to rhTPO treatment, with 10 achieving complete responses (platelet count >100 × 109/L). These findings suggest that rhTPO is a safe and effective treatment option for P-ITP that can effectively increase platelet counts and reduce the occurrence of bleeding events.
Of the four cases treated with rhTPO during pregnancy in our center, two (50%) responded to treatment, which is relatively similar to the response rate of a previous study (response rate = 74.2%). 37 Patient 1 was the first patient in our center to receive rhTPO treatment. She was treated with rhTPO at 24 weeks of gestation because of ineffective hormone and intravenous IVIg treatments, as well as a dependence on repeated platelet transfusions. Although the increase in platelet count (from 8 to 14 × 109/L) was not significant, there were no adverse events such as thrombosis during the rhTPO treatment, and she had good gestational outcomes. These findings suggest that rhTPO may be relatively safe during pregnancy. For Patient 2, rhTPO treatment showed some efficacy, and platelet counts increased from 9 to 32 × 109/L (a safe level) after treatment with 300 U/kg/day of rhTPO for 7 days. Although Patient 2 had a slight decrease in platelet counts at week 32, her counts remained at a relatively safe level (around 20 × 109/L) without any treatment in late pregnancy, demonstrating a maintenance effect of rhTPO. Patient 3 received 300 U/kg/day of rhTPO for 7 days at 29 weeks of gestation. This treatment led to a slight increase in platelet counts (from 8 to 10 × 109/L). Unfortunately, however, the curative effect was poor. In 2017, new research was published by Kong et al. 37 With the informed consent of Patient 4 and with reference to the data provided by Kong et al., we extended the treatment time of rhTPO to 14 days, with a dose not exceeding 300 U/kg/day. Surprisingly, Patient 4 exhibited very good curative effects, with platelet counts increasing from 12 to above 50 × 109/L and remaining at this level until delivery.
The cases in our center can be compared with those of Kong et al. 37 In both studies, all cases were pregnant at more than 12 weeks of gestation. Moreover, the pregnant patients had poor responses to hormone and/or IVIg first-line treatments and were dependent on blood transfusions. Platelet counts were at extremely low levels when the patients were treated with rhTPO, and some patients also had skin and mucosal bleeding. The baseline platelet counts of the patients were also comparable. However, the treatment patterns for Patients 1 to 3 differed from those of the study by Kong et al. Compared with the 14-day treatment method of Kong et al., the duration in the present study was only 7 days for Patients 1 to 3. Moreover, our treatment of the first three patients was not ideal. Changing the type of glucocorticoid was likely a reason for the poor responses to corticosteroids in our cases. In addition, the Kong et al. study provided maintenance treatment from the late stage of pregnancy to postpartum, which differs from our study. This is because rhTPO exhibited moderate treatment effects in our cases, and no patients achieved complete responses (platelet counts > 100 × 109/L or twice the baseline level), which resulted in insufficient confidence for maintaining treatment. Finally, Kong et al. treated 24 of their patients (77.4%) before 28 weeks of gestation, whereas the four cases in the present study were treated with rhTPO at 24, 27, 29, and 29 weeks of gestation. Our patients were therefore treated in relatively late stages of pregnancy, with minor therapeutic effects in maintenance treatment. This difference occurred because we aimed to ensure that the patients maintained safe levels of platelet counts until successful delivery was achieved.
Together, our findings suggest that patients with P-ITP can benefit from rhTPO treatment. Aside from rhTPO, TPO-RAs (e.g., romiplostim, eltrombopag, and avatrombopag) also show efficacy in refractory P-ITP.38–42 In a multicenter study, 40 15 women with ITP were treated with eltrombopag or romiplostim during 17 pregnancies, with the successful delivery of 18 neonates (one twin pregnancy). These findings indicate that the transient use of TPO-RAs in pregnant women with refractory ITP (refractory to at least corticosteroids and IVIg) is safe. However, the safety of TPO-RAs in pregnancy remains debatable. For example, it has been reported that a 27-year-old woman in her first pregnancy was diagnosed with severe ITP at 28 weeks of gestation. After high-dose hormone and intravenous IVIg treatment as well as repeated blood transfusion, her platelet count remained low. Moreover, the patient’s condition was complicated with gestational diabetes, which was unable to be controlled with high-dose insulin. She was administered romiplostim at 33 weeks of gestation. At 4 days of age, her neonate developed severe idiopathic thrombocytopenia purpura and grade 1 cerebral hemorrhage, requiring IVIg and platelets. 43 TPO-RAs are small molecules that easily penetrate the placental barrier in pregnant patients, and have some toxicity. Although an early study reported that thrombogenic drugs (e.g., rhTPO, romiplostim, and eltrombopag) carry a risk of thrombosis, 44 recent studies have not revealed any increase in thrombotic events.45,46 Overall, rhTPO may therefore be an alternative treatment for patients with refractory P-ITP for whom first-line treatment fails.
Conclusion
Despite the apparent limitations of the present study, our findings indicate that rhTPO treatment may be safe for patients with refractory P-ITP, and that subcutaneous injection is a convenient delivery method without adverse events (e.g., thrombosis). A longer continuous treatment may result in greater benefits to patients. Ongoing clinical studies, including ITP-PKU2204, are currently investigating the safety of rhTPO use in pregnant women. These studies aim to provide valuable insights into the potential benefits and risks of rhTPO treatment for pregnant individuals with ITP. More relevant data are needed to provide reliable treatment indications for refractory P-ITP.
Footnotes
Author contributions
JY conceived and generally supervised the study and drafted the manuscript. JY and PM collected the clinical data. SQ helped to revise the manuscript and gave final approval for the manuscript. All authors have read and approved the final manuscript.
Declaration of conflicting interest
The authors declare that there is no conflict of interest.
Ethics statement
Written informed consent to publish this case report was obtained from each patient. Because all cases were identified during routine diagnostics, ethical approval and consent to participate were not required (other than consent to undergo diagnostic testing to solve the clinical question). All patients consented to their treatments, and all patient details were de-identified.
Funding
This study was supported by the Hangzhou Technology Project of Medicine (grant number Z20210039) and the Hangzhou Major Project of Technology (grant number 202004A15).
