Journal of Thrombosis and Thrombolysis

, Volume 37, Issue 3, pp 279–286

Prednisone versus high-dose dexamethasone for untreated primary immune thrombocytopenia. A retrospective study of the Japan Hematology & Oncology Clinical Study Group

Authors

  • Kana Sakamoto
    • Division of HematologySaitama Medical Center, Jichi Medical University
  • Hideki Nakasone
    • Division of HematologySaitama Medical Center, Jichi Medical University
  • Shigeharu Tsurumi
    • Department of Hematology and OncologyDokkyo Medical University
  • Ko Sasaki
    • Department of Hematology and OncologyDokkyo Medical University
  • Kinuko Mitani
    • Department of Hematology and OncologyDokkyo Medical University
  • Michiko Kida
    • Division of HematologyNTT Kanto Medical Center
  • Akira Hangaishi
    • Division of HematologyNTT Kanto Medical Center
  • Kensuke Usuki
    • Division of HematologyNTT Kanto Medical Center
  • Ayako Kobayashi
    • Division of Hematology, Department of Internal MedicineNational Defense Medical Collage
  • Ken Sato
    • Division of Hematology, Department of Internal MedicineNational Defense Medical Collage
  • Mariko Karasawa-Yamaguchi
    • Department of HematologyToranomon Hospital
  • Koji Izutsu
    • Department of HematologyToranomon Hospital
  • Yasushi Okoshi
    • Department of HematologyUniversity of Tsukuba
  • Shigeru Chiba
    • Department of HematologyUniversity of Tsukuba
    • Division of HematologySaitama Medical Center, Jichi Medical University
Article

DOI: 10.1007/s11239-013-0939-3

Cite this article as:
Sakamoto, K., Nakasone, H., Tsurumi, S. et al. J Thromb Thrombolysis (2014) 37: 279. doi:10.1007/s11239-013-0939-3

Abstract

High-dose dexamethasone (HDD) has been shown to be an effective initial treatment for immune thrombocytopenia (ITP), but it is not clear whether HDD offers any advantages over conventional-dose prednisone (PSL). We retrospectively compared the efficacy and toxicity of HDD and PSL for newly diagnosed ITP. The response was evaluated according to the International Working Group (IWG) criteria. We analyzed data from 31 and 69 patients in the HDD and PSL groups, respectively. There were no significant differences in patient characteristics between the two groups except for the incidence of the eradication of Helicobacter pylori. The response rate was better in the HDD group (42.7 vs. 28.4 %), and this difference was statistically significant when adjusted for other factors including the eradication of H. pylori. In the HDD group, a response was achieved earlier (28 vs. 152 days in median) and steroids were more frequently discontinued at 6 months (64.5 vs. 37.7 %). Among patients who achieved a response, there was no significant difference in the incidence of loss of response. There were no significant differences in the rate of adverse events, transition to chronic ITP, and splenectomy. In conclusion, HDD might enable the early cessation of steroids without a loss of response.

Keywords

Immune thrombocytopeniaPrednisoneHigh dose dexamethasoneResponse rateClinical management

Introduction

Immune thrombocytopenia (ITP) is an immune-mediated acquired disease that is characterized by a transient or persistent isolated decrease in the platelet count and an increased risk of bleeding [1]. Although the pathogenesis of ITP is not yet fully understood, the antibody- and cell-mediated destruction of platelets and the suppression of platelet production are believed to be involved. According to the IWG criteria [2], primary ITP is defined as an autoimmune disorder that is characterized by isolated thrombocytopenia (peripheral blood platelet count of <100 × 109/L) in the absence of other causes or disorders that may be associated with thrombocytopenia.

Corticosteroids have been widely recognized as the most appropriate first-line treatment [3, 4]. Conventionally, prednisone (PSL) is chosen as a first-line therapy for adult patients with ITP who require treatment. The starting dose of 0.5–1 mg/kg/day is usually continued for 2 to 4 weeks and then gradually tapered over several weeks. According to previous reports, an initial response was achieved in 50–60 % of patients by this approach with PSL, whereas the long-term remission rate was <30 % [510]. High-dose dexamethasone (HDD) has been shown to be an effective initial treatment for ITP, with a sustained response in 36–75 % of patients [1114]. However, its advantage over conventional-dose PSL has not been clarified. Therefore, we retrospectively compared the efficacy and toxicity of HDD and PSL for newly diagnosed ITP based on the new standardized criteria from the IWG [2].

Patients and methods

We collected data from adult newly diagnosed primary ITP patients with platelet counts of <30 × 109/L, who were initially treated with PSL or HDD at six institutions that participated in the Japan Hematology and Oncology Clinical Study Group (J-HOCS) between January 2005 and December 2009 through a questionnaire survey. Basically, the initial treatment was selected according to the practical policy of the responsible institution. Recurrent or chronic ITP patients were excluded. The HDD group included both patients who were treated with HDD alone and those who were treated with HDD followed by PSL or sequential HDD. Concurrent use of intravenous immunoglobulin (IVIG) was allowed. Prior or concurrent eradication of Helicobacter pylori was also allowed. Bleeding symptoms were graded on a scale from 0 to 4 according to a scoring system reported elsewhere [12]. Grade 0 was defined as no bleeding symptoms and grade 4 was defined as major hemorrhage with sequelae and/or a life-threatening event or death. Bleeding symptoms were considered to be present if the bleeding score was graded 1 or higher.

The response was evaluated according to the IWG criteria [2]: complete response (CR) was defined as any platelet count of at least 100 × 109/L; response (R) was defined as any platelet count between 30 and 100 × 109/L and at least double the baseline count; and no response (NR) was defined as any platelet count <30 × 109/L or less than double the baseline count. Steroid dependence was defined as the need for ongoing or repeated administration of corticosteroids for at least 2 months to maintain a platelet count of at least 30 × 109/L. Patients with corticosteroid dependence were considered non-responders. Therefore, patients who achieved a platelet-count response but continued to receive steroids for at least 2 months were excluded from the responders. Loss of CR was defined as a decrease in the platelet count to below 100 × 109/L or bleeding, and loss of R was defined as a decrease in the platelet count to below 30 × 109/L or less than a twofold increase from the baseline platelet count or bleeding. Chronic ITP was defined as ITP that lasted for >12 months from the diagnosis.

HDD could be repeatedly administered as first-line therapy, but only HDD that was administered within 2 months from the initiation of first-line therapy was considered first-line therapy. Patients who required second-line therapy were treated as non-responder. Second-line therapy was defined as the resumption or dose-escalation of steroids beyond 2 months after the start of the initial therapy, any medications other than steroids, and splenectomy. For example, if HDD was administered every 2 weeks for six cycles, the 5th cycle was considered to be second-line treatment. Eradication of H. pylori that occurred beyond 2 months after the initiation of first-line therapy was also considered second-line therapy.

Statistical analysis

Days to these events were calculated from the date of the initiation of the first-line therapy. Comparisons between the two groups were performed with Fisher’s exact test for categorical variables and the Mann–Whitney U-test for continuous variables. We assessed cumulative incidences of R and CR using Gray’s methods while considering death without R or CR and the transition to secondary treatment as competing risks. We used Gray’s methods because this is the test designed for comparing the cumulative incidence curves among different groups in the presence of competing risks [15, 16]. In a multivariate analysis using Fine-Gray’s proportional hazards model, the effect of first-line therapy was adjusted for the eradication of H. pylori, which was significantly more often used in the HDD group, and the other factors with p value of less than 0.15 by a univariate analysis, which were deleted from the model in a stepwise manner if they were not significant in the multivariate analysis. Estimates of relative risk (RR) and corresponding 95 % confidence intervals (95 %CI) were also obtained for each variable.

Statistical significance was defined as a 2-sided p value of less than 0.05. All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University, at http://www.jichi.ac.jp/saitama-sct/SaitamaHP.files/statmedEN.html), which is a graphical user interface for R (The R Foundation for Statistical Computing, version 2.13.0). More precisely, it is a modified version of R commander (version 1.6-3) that was designed to add statistical functions that are frequently used in biostatistics [17]. This study was approved by the institutional review board of Jichi Medical University.

Results

Patient characteristics

We collected data from 31 and 69 patients in the HDD and PSL groups, respectively. The overall median age was 58 years. The starting dose of PSL in the PSL group was 1 mg/kg in 31 cases, 0.5 mg/kg in 30, 0.6 mg/kg in 2, and other in 4. In the HDD group, dexamethasone at 40 mg/day for 4 days was administered only once in 11 cases, twice in 11, three times in 5, and five times in 1 within 2 months after the start of therapy. Two and one patients received dexamethasone at 20 mg/day for 4 days once and twice, respectively. Fourteen patients in the HDD group received subsequent PSL at a median dose of 0.4 mg/kg/day (range 0.25–1.0 mg/kg/day) after HDD within 2 months from the initiation of the first-line therapy.

There were no significant differences in the patient characteristics between the two groups, such as age, sex, bleeding score, comorbidity, H. pylori infection, platelet count at the initiation of therapy, or duration between diagnosis and treatment (Table 1). However, H. pylori was eradicated more frequently in the HDD group (52 vs. 26 %, p = 0.039). This eradication was mostly performed within 1 week before or after the initiation of the first-line therapy (12 cases in the HDD group and ten in the PSL group). In the other cases, the timing of eradication was as follows; in the second week after the initial therapy in two cases in the HDD group and one case in the PSL group, later than 2 weeks but within 2 months after initiation in two cases in the PSL group, and earlier than 1 week before initiation in two cases in the HDD group and five in the PSL group.
Table 1

Patient characteristics

 

HDD

PSL

Total

p value

Number

31

69

100

 

Median age (range)

55 (18–86)

61 (18–91)

58.5

0.41

Gender

 Male/Female

16/15

30/39

47/54

0.52

Median platelet count at diagnosis (× 10^9/L)

8

10

8.5

0.83

Bleeding at diagnosis

27

50

77

0.28

Bleeding score

   

0.43

 1

7

15

22

 

 2

15

25

40

 

 3

4

8

12

 

 4

1

0

1

 

Preceding infection

6

9

15

0.55

Helicobacter pylori infection

16

28

44

0.65

Eradication of H. pylori

16

18

34

0.039

Comorbidities at diagnosis

15

30

45

0.83

 Diabetes

2

4

6

1

 Hyperlipidemia

2

5

7

1

 Cardiovascular disease

2

2

4

0.59

 Hypertension

4

8

12

1

Autoimmune disease

4

6

10

0.72

Pregnancy at diagnosis

0

2

2

1

Median days from diagnosis to treatment (range)

0 (0–51)

0 (0–79)

0 (0–79)

0.48

Platelet count at initiation of treatment (× 10^9/L)

8

10

8.5

0.97

Emergent platelet transfusion

6

9

15

0.55

Intravenous immunoglobulin

4

6

10

0.50

Response

All of the patients in the HDD group and 66 of the 69 cases in the PSL group achieved a platelet count response to at least 30 × 109/L. Twenty-eight of the 31 patients in the HDD groups and 63 of the 69 in the PSL group achieved a complete platelet count response to at least 100 × 109/L. However, this response was lost in some cases because of the discontinuation of treatment or even with continuous treatment. The 1-year cumulative rate of achievement of R was 42.7 % in the HDD group and 28.4 % in the PSL group (p = 0.077, Fig. 1). Furthermore, a response was achieved earlier in the HDD group. The median duration from the start of the initial therapy to the achievement of a response was 28 and 152 days in the HDD and PSL groups, respectively (p = 0.011). Multivariate analyses revealed that HDD therapy was significantly associated with a higher response rate after adjusting for the coexistence of autoimmune disease, eradication of H. pylori, and preceding infection (p = 0.042, Table 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs11239-013-0939-3/MediaObjects/11239_2013_939_Fig1_HTML.gif
Fig. 1

Cumulative incidence of response grouped according to the initial treatment. HDD high-dose dexamethasone, PSL prednisone

Table 2

Prognostic factors for achievement of Response

Prognostic factors

Response rate (%)

Univariate p value

Multivariate hazard ratio

p value

Autoimmune disease

 

0.134

  

 Present

0.0

   

 Absent

36.8

   

Initial treatment

 

0.077

  

 HDD

42.7

 

1

0.042

 PSL

28.4

 

0.445 (0.204–0.970)

 

Eradication of Helicobacter pylori

 

0.46

  

 Present

24.1

 

0.604 (0.278–1.315)

0.200

 Absent

38.6

 

1

 

Preceding infection

 

0.149

  

 Present

41.3

   

 Absent

31.9

   

Helicobacter pylori was eradicated significantly more often in the HDD group. Therefore, we compared the R rate in patients in whom H. pylori had been eradicated to that in patients who had not undergone such eradication in the HDD group to exclude the effect of previous or concurrent eradication of H. pylori. The 6-month cumulative achievement of R tended to be lower in patients in whom H. pylori was eradicated (25.0 vs 46.7 %, p = 0.152).

The cumulative achievement of CR tended to be higher in the HDD group, but this difference was not significant (32.3 % for the HDD group and 25.1 % for the PSL group, p = 0.23) (Fig. 2). However, CR was achieved significantly earlier in the HDD group (37 days for the HDD group, and 168 days for the PSL group, p = 0.010). The multivariate analysis did not identify any factor that significantly predicted the achievement of CR.
https://static-content.springer.com/image/art%3A10.1007%2Fs11239-013-0939-3/MediaObjects/11239_2013_939_Fig2_HTML.gif
Fig. 2

Cumulative incidence of complete response grouped according to the initial treatment. HDD high-dose dexamethasone, PSL prednisone

In the HDD group, 20 of 31 patients were in remission without steroid at 6 months from the initiation of the first-line treatment. On the other hand, only 26 of the 69 in the PSL group were not receiving steroids at 6 months (p = 0.017, Fig. 3). Among patients who were followed-up for at least 1 year, chronic ITP developed in 10 of 20 (50 %) and 35 of 54 (65 %) in the HDD and PSL groups, respectively (p = 0.29).
https://static-content.springer.com/image/art%3A10.1007%2Fs11239-013-0939-3/MediaObjects/11239_2013_939_Fig3_HTML.gif
Fig. 3

Proportion of patients who were in remission without steroids among those who had been followed-up until each month in the corresponding group

Loss of response and secondary treatments

There was no difference in the incidence of the loss of response between the two groups. A loss of R was observed in 3 of 13 and 1 of 13 of the patients who achieved R in the HDD and PSL groups, respectively (p = 0.31). A loss of CR was observed in 4 of 12 and 2 of 12 of the patients who achieved CR in the HDD and PSL groups, respectively (p = 0.51).

Various secondary treatments including corticosteroids, splenectomy, and other medications were applied in 14 patients (45 %) in the HDD group and 18 (26 %) in the PSL group (p = 0.068, Table 3). These secondary treatments tended to be performed more frequently and were started earlier in the HDD group (day 78 for the HDD group and day 148 for the PSL group, p = 0.014). Splenectomy was performed in seven patients in the HDD group and six in the PSL group (p = 0.10).
Table 3

Secondary treatment

Secondary treatment

HDD

PSL

HDD (+PSL)

3

1

PSL

3

3

Splenectomy

7

6

Cepharanthine

0

4

Eradication of Helicobacter pylori

0

3

Danazol

2

0

Cyclosporine

1

1

Total

14

18

In the HDD group, one patient received cyclosporine and splenectomy and another patient received danazol and splenectomy

Adverse events and mortality

As adverse events, we evaluated the incidences of frequent side effects of corticosteroids, such as diabetes, hyperlipidemia, hypertension, compression fracture, infection, peptic ulcer, and psychotic symptoms, but found no significant differences between the groups (Table 4). Documented infection tended to be observed more frequently in the HDD group (5 and four cases in the HDD and PSL groups, respectively. p = 0.13). In the HDD group, oral candidiasis was observed in 2, upper respiratory infection in 1, enteritis in 1, and peritonitis after splenectomy in 1. In the PSL group, Herpes zoster virus reactivation was seen in 2, urinary tract infection in 2, Pneumocystis jiroveci pneumonia in 1, and retroperitoneal abscess in 1 after splenectomy.
Table 4

Adverse events

 

HDD (n = 31)

PSL (n = 69)

p value

Exacerbation of Diabetes

1

3

1

Steroid Diabetes

3

3

0.37

Hyperlipidemia

3

5

0.7

Hypertension

0

0

Avascular necrosis of the femoral head

0

0

Compression fracture

1

0

0.31

Congestive heart failure

0

0

Documented infection

5

4

0.13

Peptic ulcer

0

0

Psychotic symptoms

1

4

1

There was one death in each group during follow-up. There was no significant difference in the mortality rate between the two groups (3.2 % in HDD, 1.4 % in PSL, p = 0.52). A 45-year-old female in the HDD group died of cerebral hemorrhage on day 57, and an 88-year-old male in the PSL group died of cerebral hemorrhage on day 3. For the patient in the HDD group, the platelet count on the day of the event was 5 × 109/L and that of the day she died was 3 × 109/L. For the patient in the PSL group, the platelet count on the day of cerebral hemorrhage was not checked, but it was 14 × 109/L 2 days before. Since this patient was 88 years old and his family did not want further examination after the cerebral hemorrhage, the platelet count was not checked thereafter. No infectious death was observed during follow-up.

Discussion

Corticosteroids have been widely recognized as the most appropriate initial treatment for ITP patients, but the response rate remained 50–60 % with prednisone at a starting dose of 0.5–1 mg/kg/day and the long-term remission is obtained in only 5–30 % of patients after the reduction or discontinuation of the therapy [510]. Therefore, long-term corticosteroid therapy is required to maintain the platelet count and a considerable number of patients are suffered from side effects including serious infections [10, 18]. In this context, based on the efficacy of pulsed corticosteroids in reducing immunoglobulin production in clonal B cell disorders like multiple myeloma and long half-life and good tolerability of dexamethasone in high doses, HDD was tested and showed an excellent outcome in a small cohort of patients with refractory ITP [19]. In early-2000s, Cheng et al. [11] applied HDD as initial treatment for untreated ITP and showed a good response rate of 85 % and a sustained response of 50 % with 30.5 months of follow-up. Borst et al. [13] also showed similar results with repeated courses of HDD. Recently, Mazzucconi et al. conducted a multicenter study with modified HDD regimen for untreated ITP patients and showed a response rate of 85.6 % and a persistent response of 74 % [12]. In these studies, the treatment was well tolerated. Furthermore, it should be noted that the response occurred early in the course with HDD compared to the conventional corticosteroid therapy which takes 7–10 days mostly. In the study of Cheng et al. the platelet count of the responders had increased by at least 20 × 109/L by the third day of treatment and the mean count was 101 × 109/L 1 week after the initiation of treatment. Mazzucconi et al. also showed that the median count of platelets was already 87 × 109/L at fourth day of the first therapy. The early attainment of response could reduce serious and fatal hemorrhage in the early phase of the disease.

We retrospectively analyzed the efficacy and toxicity of HDD and PSL for newly diagnosed ITP. The response rate was better in the HDD group, with cumulative achievements of R of 42.7 and 28.4 % in the HDD and PSL groups, respectively, and this difference was significant by a multivariate analysis. Moreover, the response was achieved earlier and a steroid-free status at 6 months was achieved more often without a loss of response in the HDD group. These results were compatible with data from previous studies which showed sustained response rates of 36 to 75 % with HDD and 5 to 30 % with conventional-dose PSL [514]. With regard to adverse events, there was no significant difference in the rate of adverse events between the two groups, even though the patient characteristics, such as age, gender, and initial platelet count, were similar.

There was no significant difference in the patient characteristics between the two groups except that the rate of prior or concurrent eradication of H. pylori was significantly higher in the HDD group. However, in the HDD group, the R rate in patients who underwent the eradication of H. pylori tended to be lower than that in patients who did not. In addition, the response rate was significantly higher in the HDD group even when we excluded patients who underwent prior or concurrent eradication of H. pylori (53.3 vs. 26.7 %, p = 0.0056). Therefore, the better response in the HDD group compared to the PSL group was not due to the more frequent use of prior or concurrent eradication of H. pylori.

The transition to secondary treatments occurred earlier in the HDD group in this study, although there was no significant difference in the rate of transition to secondary treatments. This may have been because the generally faster achievement of R in the HDD group enabled physicians to make an earlier decision to proceed to secondary therapy. Novel salvage treatments for refractory ITP, such as rituximab [14, 2023] or thrombopoietin receptor agonists [2430], have recently become available, and therefore an earlier decision to proceed to secondary treatments would be beneficial for avoiding unnecessary adverse events due to steroids.

This study was limited due to its retrospective nature and small sample size although eligible patients were included consecutively in the study. There might be a selection bias between the HDD and PSL groups, since a physician could avoid HDD as an initial treatment for a possible intolerant patient. However, the choice of the initial treatment largely depended on the institution’s policy rather than the individuality of each patient. Some institutions changed their first-line therapy from PSL to HDD at a certain point and others consistently use PSL as first-line therapy. Therefore, the findings of this study were minimally affected by the selection bias. Moreover, there were no significant differences in the patient characteristics between the HDD group and the PSL group other than the rate of eradication of the H. pylori. In addition, the dose and duration of corticosteroids were heterogeneous in both groups. A previous study showed that patients who received three or more courses of HDD responded better than those who received one or two courses of HDD [12]. The response rate in the HDD group in our cohort may have been underestimated, since patients in the HDD group in this study mostly received less than three courses.

In conclusion, we retrospectively compared the efficacy and adverse effects of HDD and PSL for newly diagnosed ITP. The response rate was better and the response was achieved earlier in the HDD group. These findings suggest that HDD might enable the early cessation of steroids without a loss of response. However, these findings should be confirmed in a randomized controlled trial.

Copyright information

© Springer Science+Business Media New York 2013