Articles
Rituximab as second-line treatment for adult immune thrombocytopenia (the RITP trial): a multicentre, randomised, double-blind, placebo-controlled trial Waleed Ghanima, Abderrahim Khelif, Anders Waage, Marc Michel, Geir E Tjønnfjord, Neila Ben Romdhan, Johannes Kahrs, Bernadette Darne, Pål Andrè Holme, on behalf of the RITP study group*
Summary Background Immune thrombocytopenia is characterised by immune-mediated destruction and suboptimum production of platelets. Despite the absence of supporting evidence, rituximab is frequently used off-label in patients with immune thrombocytopenia. We aimed to assess the efficacy of rituximab as compared with placebo as a splenectomy-sparing treatment in patients who were previously treated with corticosteroids. Methods In this multicentre, randomised, double-masked, placebo-controlled trial, we enrolled corticosteroid unresponsive adult patients (aged ≥18 years) with primary immune thrombocytopenia and a platelet count of less than 30 × 10⁹ platelets per L. Patients were randomly assigned (1:1) to four weekly infusions of 375 mg/m² rituximab or placebo. Concurrent treatment with corticosteroids only was allowed during the study. The primary endpoint was rate of treatment failure within 78 weeks—a composite of splenectomy or meeting criteria for splenectomy after week 12 if splenectomy was not done, assessed in all patients who received at least one dose of study treatment. Secondary endpoints were response rates, relapse rates, and duration of response. Efficacy endpoints were assessed with the Kaplan-Meier method. Safety endpoints were assessed in all patients who received at least one dose. This trial is registered with ClinicalTrials.gov, number NCT00344149. Findings Between Aug 17, 2006, and June 30, 2011, we enrolled 112 patients. 32 (58%) of 55 patients in the rituximab group and 37 (69%) of 54 patients in the placebo group had treatment failure within 78 weeks (Kaplan-Meier cumulative incidence 46% for rituximab vs 52% for placebo (hazard ratio [HR] 0·89, 95% CI 0·55–1·45; p=0·65). The cumulative incidence of overall response was 81% in the rituximab group versus 73% in the placebo group (p=0·15) and complete response was 58% in the rituximab group versus 50% in the placebo group (p=0·12). Of those achieving an overall response, 68% relapsed in the rituximab group and 78% relapsed in the placebo group, and of those achieving complete response, 50% relapsed in the rituximab group and 62% relapsed in the placebo group. Time to relapse in the rituximab group was longer in patients who achieved overall response (36 vs 7 weeks; p=0·01) but not complete response (76 vs 49 weeks; p=0·19). Rates of bleeding were similar in the two groups (21 [38%] in the rituximab group vs 27 [50%] in the placebo group; p=0·08) as were rates of infection (22 [40%] vs 13 [24%]; p=0·09). Interpretation Despite no reduction in the rate of long-term treatment failure with rituximab, a small benefit cannot be ruled out, as suggested by an apparently longer duration of response and numerically higher response rates with rituximab. Funding South-East Regional Health Authority and Østfold Hospital, Norway; Roche, France; and Roche, Norway.
Introduction Immune thrombocytopenia is caused by immunemediated platelet destruction and impaired platelet production, causing thrombocytopenia and occasionally bleeding.1 The main goals of treatment are: first, to achieve sustained, haemostatic platelet counts to minimise the risk of bleeding; and second, to cure the patient if possible—both aims should be achieved with the fewest possible toxic effects.2,3 Corticosteroids are the main first-line treatment for immune thrombocytopenia in adults, yielding initial response rates of 70–90%;2,3 however, because of relapses during dose tapering or after corticosteroid discontinuation, sustained responses are reduced and occur between 30% and 80% of the time.2,4–6 Patients who do not achieve a sustained
response to corticosteroids usually need second-line treatment. Splenectomy is still recommended as the standard second-line therapy in patients with chronic immune thrombocytopenia.3 Splenectomy is an invasive procedure associated with perioperative and postoperative complications,7,8 with no means to predict the patient’s response to the procedure.9 Because of the availability of medical alternatives2 and increasing evidence of late remissions, whether spontaneous or with continuing treatment,5,10 physicians have become more reluctant to do a splenectomy and patients are less willing to undergo splenectomy.9,11 Of the many available second-line alternatives,2 the anti-CD20 antibody rituximab and approved thrombopoietin receptor agonists are widely used.9 Rituximab has
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
Published Online February 5, 2015 http://dx.doi.org/10.1016/ S0140-6736(14)61495-1 See Online/Comment http://dx.doi.org/10.1016/ S0140-6736(14)61930-9 *Members listed at end of paper Department of Medicine, Østfold Hospital Trust, Fredrikstad, Norway (W Ghanima MD, J Kahrs MD); Department of Medicine, Farhat Hached Hospital, Sousse, Tunisia (Prof A Khelif MD); Department of Medicine, St Olav Hospital, Trondheim, Norway (Prof A Waage MD); Department of Medicine, Henri Mondor Hospital, Université Paris-Est Creteil, Créteil, France (Prof M Michel MD); Department of Medicine, La Rabta Hospital, Tunis, Tunisia (N B Romdhan MD); Haematology Department, Oslo University Hospital and Institute of Clinical Medicine, Oslo University, Oslo, Norway (W Ghanima, Prof G E Tjønnfjord MD, Prof P A Holme MD); and Monitoring Force, MaisonsLaffitte, France (B Darne MD) Correspondence to: Dr Waleed Ghanima, Department of Medicine, Østfold Hospital Trust, 1603 Fredrikstad, Norway
[email protected]
1
Articles
been shown to provide initial response rates of about 60%;12 however, sustained responses lasting for more than 2 years occur in substantially fewer than 60% of patients.13,14 Whether the administration of rituximab to patients who did not respond to corticosteroid treatment would render splenectomy or other therapeutic options superfluous is unknown. This trial aimed to assess the long-term efficacy of rituximab as a splenectomy sparing option in corticosteroid-exposed adult patients with immune thrombocytopenia.
Methods Study design and patients In this double-masked, placebo-controlled trial, we enrolled adult patients with immune thrombocytopenia10 in 14 centres in Norway, Tunisia, and France. The eligibility criteria for inclusion in the study were unsplenectomised patients with primary immune thrombocytopenia with a platelet count of less than 30 × 10⁹ platelets per L (or 30–50 × 10⁹ platelets per L if a higher platelet count was deemed necessary because of bleeding or concomitant anti-platelet therapy); patients who did not achieve a sustained response to 1–2 mg/kg prednisone or prednisolone given for a minimum duration of 2 weeks, or relapse during steroid-tapering or after its discontinuation; and aged 18 years or older. The main exclusion criteria were previous administration of any second-line treatment for immune thrombocytopenia; immune thrombocytopenia secondary to systemic lupus erythematosus or haematological malignancies; pregnancy or breastfeeding; evidence of HIV or hepatitis B or C infections; heart failure (New York Heart Association class higher than 2); an underlying malignancy; expected survival of less than 2 years; known primary or secondary immune deficiency syndromes; severe chronic pulmonary obstructive disease; and acute coronary syndrome during the past 6 months. The 112 patients randomly assigned
58 assigned to rituximab 55 received allocated intervention 3 did not receive allocated intervention because treatment refused
54 assigned to placebo 54 received allocated intervention 0 did not receive allocated intervention
55 followed up 0 lost to follow-up 0 patient refusal to continue the study 1 change in diagnosis
52 followed up 1 lost to follow-up 1 patient refusal to continue the study 0 change in diagnosis
55 included in the analysis 0 excluded from analysis
54 included in the analysis 0 excluded from analysis
Figure 1: Trial profile
2
study was approved by the ethics committees in all countries. All patients provided written informed consent before enrolment.
Randomisation and masking We randomly assigned patients in a 1:1 ratio to receive rituximab or placebo. Randomisation was centralised at Østfold Hospital, Norway with precoded concealed envelopes. We used permuted block randomisation with stratification on the centre with a block size of four. Treatment allocation was communicated to local hospital pharmacies by fax. Infusion bags were concealed in opaque plastic bags that were sealed to mask the solution from the patient and health-care providers. Pharmacists were not masked to treatment allocation. Patients, investigators, and reviewers were masked to allocation.
Procedures Eligible patients received four weekly infusions of 375 mg/m² rituximab (Mabthera; Roche, Basel, Switzerland) or infusions of NaCl 0·9% as placebo. Premedication to reduce infusion-related reactions included oral acetaminophen and intravenous antihistamine. Patients were allowed to continue taking corticosteroids after randomisation. The steroid dose had to be tapered gradually to the lowest dose that kept the platelet count higher than 20 × 10⁹ platelets per L. Study visits were scheduled every 6 weeks during the study for 78 weeks (1·5 years) or for 12 weeks after splenectomy. At each visit, we recorded platelet counts, bleeding score, steroid-dose, platelet increasing or concomitant medication, and adverse events.
Outcomes The primary endpoint was amended from splenectomy within 78 weeks to the rate of treatment failure, defined as a splenectomy or meeting the criteria for a splenectomy after week 12, if a splenectomy was not done because of contraindications or patients’ refusal. The amended protocol was approved by the steering committee in October, 2010, and the final version approved on Jan 18, 2011. The primary endpoint was amended to treatment failure or splenectomy because thrombopoietin receptor agonists (romiplostim and eltrombopag) were approved, and it became unethical to deny patients with immune thrombocytopenia this therapeutic option; and because a splenectomy was occasionally not done because of patients’ refusal or because the presence of contraindication. The criteria for splenectomy were defined as a platelet count lower than 20 × 10⁹ platelets per L or a need for prednisone or prednisolone increments at doses higher than 7·5 mg/day to maintain a platelet count of less than or equal to 20 × 10⁹ platelets per L. Secondary endpoints were response rates, duration of response, safety, bleeding, and corticosteroid consumption. Response rates were defined as platelet
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
Articles
count of at least 30 × 10⁹ platelets per L (overall response) or at least 100 × 10⁹ platelets per L (complete response), after week 4 from first study drug administration. At least a doubling of the platelet count from baseline without administration of any platelet increasing therapy, except stable or decreasing doses of prednisone or prednisolone, during the past 4 weeks was also needed to qualify as a response. Relapse was defined as platelet counts lower than 30 × 10⁹ platelets per L after achieving a response. Response duration was defined as time from achieving an overall response or complete response to relapse. Bleeding, infections, and thrombosis were the main safety endpoints. Toxic effects were graded according to Common Terminology Criteria for Adverse Events (CTCAE) version 3.0.15 Bleeding was graded according to the WHO criteria for bleeding into five grades: 0, no bleeding; 1, petechiae; 2, mild blood loss; 3, gross blood loss; and 4, debilitating blood loss. All corticosteroids were converted to doses equivalent to prednisolone.
Statistical analysis We hypothesised that rituximab would reduce the rate of treatment failure within 78 weeks from 70% to 40%— ie, an absolute risk reduction of 30%, corresponding to a hazard ratio of rituximab to placebo of 2·57. With a two-sided log-rank test, power of 80%, and type 1 error of 5%, 41 events were needed, corresponding to 46 patients randomly allocated to each group. With the assumption that 20% of patients would drop out before treatment failure, the final sample size was estimated to be 110. We assessed outcomes in a full analysis set of all randomly allocated patients who received at least one dose of study drug. The Kaplan-Meier method and Rituximab (n=55)
Placebo (n=54)
Median age, years
46 (27–61)
46 (28–60)
Female
40 (73%)
39 (72%)
Median platelet count, ×10⁹ cells/L
16 (6–27)
21 (9–29)
Median bleeding score*
2
4
6 (11%)
7 (13%)
Mild blood loss
19 (35%)
23 (43%)
Gross blood loss
2 (4%)
2 (4%)
Treatment with corticosteroids
32 (58%)
24 (44)
Duration of ITP in weeks, median (IQR)
37 (8–288)
50 (14–211)
Newly diagnosed (0–3 months)
18 (33%)
12 (22%)
Persistent (3–12 months)
13 (24%)
16 (30%)
Chronic (>12 months)
24 (44%)
26 (48%)
Data are median (IQR) or n (%). ITP=immune thrombocytopenia. *Bleeding was also reported by a bleeding score.16
Table 1: Baseline characteristics
Role of the funding source The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. All authors had full access to the data and had final responsibility for the decision to submit for publication.
Results Between Aug 17, 2006, and June 30, 2011, we enrolled 112 patients (figure 1, table 1). Of 112 randomly assigned patients, three patients did not receive the allocated treatment and were not included in the full analysis set. Two patients in the placebo group were lost to follow-up. Table 2 summarises the number of patients who responded to treatment and serious safety outcomes. The Kaplan-Meier estimates of cumulative incidence of overall response at 78 weeks were 81% in the rituximab group and 73% in the placebo group (p=0·15) and were 58% in the rituximab group and 50% in the placebo group for complete response Rituximab (n=55)
Placebo (n=54)
p value*
Efficacy outcomes Treatment failure
32 (58%)
37 (68%)
0·65
Splenectomy
8 (15%)
14 (26%)
0·12
Overall response
40 (73%)
36 (67%)
0·15
Loss of overall response
27 (68%)
28 (78%)
0·01
Median duration of overall response (weeks)
36 (13–not reached)
7 (5–69)
0·01
28 (51%)
21 (39%)
0·12
14 (50%)
13 (62%)
0·19
49 (20–95)
0·19
Complete response Loss of complete response
Bleeding Petechiae
log-rank test were used for efficacy outcomes. Logistic regression was used to establish the association between certain baseline parameters and response to treatment at 78 weeks. Efficacy outcomes were adjudicated by two reviewers who were masked to treatment allocation. Descriptive statistics were used to summarise demographic and baseline characteristics, and safety data. All statistical tests were two-sided with a significance of 5%. Analyses were done with SAS version 9.2. This study is registered with ClinicalTrials.gov, number NCT00344149.
Median duration of 76 (32–not complete response (weeks) reached) Main safety outcomes Death
1 (2%)
NA
Bleeding
21 (38%)
27 (50%)
0·08
Infections
22 (40%)
13 (24%)
0·09
Venous thrombosis†
0
2 (4%)
0
NA
Data are n (%) or median (IQR). NP=testing was not done for difference in death and venous thrombosis because of the small number of events. *Log-rank p value estimated from survival analysis. †One pulmonary embolism and one deep venous thrombosis.
Table 2: Number and simple rates of the efficacy and safety outcomes of the study
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
3
Articles
A Placebo Rituximab
1·0
1–Probability of an event
0·8
0·6
0·4
Log-rank p=0·1541
0·2
0 0 Number at risk Placebo Rituximab
52 55
20 47 41
40 30
27 20
40 Time to response (weeks) 19 18
16 10
14 9
14 9
60 12 9
11 8
11 8
80 11 8
11 8
11 8
B 1·0
1–Probability of an event
0·8
Log-rank p=0·1195
0·6
0·4
0·2
0 0 Number at risk Placebo Rituximab
20
40
60
80
Time to complete response (weeks) 52 55
49 45
47 36
32 26
26 25
20 15
18 14
18 14
15 14
14 14
14 13
14 13
13 13
13 12
Figure 2: Time to overall response (A) or complete response (B) within 78 weeks Overall response was defined as a platelet count of at least 30 × 10⁹ cells per L and complete response was defined as a platelet count of at least 100 × 10⁹ cells per L, after week four from first study drug administration. At least a doubling of platelet count from baseline without administration of any platelet elevating therapy except stable or decreasing dose of prednisone or prednisolone during the past 4 weeks was needed to classify overall response.
(p=0·12; figure 2). Hazard ratios were estimated using a Cox proportional hazard model. The Kaplan-Meier estimate of cumulative incidence of treatment failure was 46% in the rituximab group and 52% in the placebo group (HR 0·89, 95% CI 0·55–1·45, p=0·65; figure 3). Of these patients, eight (15%) had a 4
splenectomy in the rituximab group and 14 (26%) did so in the placebo group (p=0·12). Figure 4 shows the platelet counts in each study group. Median time to relapse in patients who achieved overall response was 36 weeks (interquartile range [IQR] 13–not reached) in the rituximab group and
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
Articles
Placebo Rituximab
1·0
1–Probability of an event
0·8 Log-rank p=0·6461 0·6
0·4
0·2
0 0 Number at risk Placebo Rituximab
54 55
20 52 55
51 54
43 48
40 Time to treatment failure (weeks) 40 48
35 41
30 37
29 35
60
28 35
27 32
80
27 31
26 30
25 30
25 28
Figure 3: Time to treatment failure within 78 weeks The composite outcome of splenectomy or meeting criteria for splenectomy after week 12 if splenectomy was not done because of contraindications or patient’s refusal.
Placebo Rituximab
210 180
Platelet count (×109 per L)
7 weeks (5–69) in the placebo group (p=0·014; figure 5), and in patients who achieved complete response, the median times to relapse were 76 weeks (31–not reached) and 49 weeks (20–95), respectively (p=0·19) (figure 5). Neither age, platelet count, duration of immune thrombocytopenia, nor body-mass index at baseline was associated with response. Three patients (3%), two in the rituximab group and one in placebo group, eventually achieved complete response without any treatment after being initially classified as not responding to treatment; they were deemed spontaneous remitters and were not included in the response analyses. One or more bleeding episodes were registered in 21 (38%) patients in the rituximab group and 27 (50%) patients in the placebo group. The first bleeding episode occurred earlier in the placebo group (p=0·08; appendix). Maximum severity of bleeding was graded as mild blood loss (grade 2) in 34 patients (15 in the rituximab group; 19 in the placebo group) and gross blood loss (grade 3) in six patients (four in the rituximab group; two in the placebo group). One patient who received the placebo died 35 weeks after randomisation because of profuse gastrointestinal bleeding (appendix). The patient was a non-responder. One or more infections were registered in 22 patients in the rituximab group and 13 in the placebo group (p=0·09; appendix). One event was graded as grade 3 in each group (appendicitis in the rituximab group and pneumonia in the placebo group). Median serum concentrations of immunoglobulin at 24 weeks and 78 weeks were close to pretreatment concentrations (table 3). Two cases of venous thromboembolism were
150 120 90 60 30 0 0
6
24
42
60
78
Time (weeks)
Figure 4: Median and IQR of platelet counts (×10⁹/L) at each study visit (every 6 weeks) in the two study groups Platelet values were censored after splenectomy.
recorded in the rituximab group; one pulmonary embolism and one deep vein thrombosis. Table 4 summarises adverse events occurring in at least 5% of patients, or events that were greater than grade 2 in severity. No significant difference was noted in the probability of receiving rescue treatment (platelet transfusion or intravenous immunoglobulin) between the two groups (p=0·09; appendix). Total cumulative doses of corticosteroids consumed during weeks 0–24 were 56 g in the rituximab group and 70 g in the placebo group, and
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
5
Articles
A Placebo Rituximab
1·0
1–Probability of an event
0·8
0·6 Log-rank p=0·0143 0·4
0·2
0 0 Number at risk Placebo Rituximab
36 40
20 21 37
15 31
11 28
40 Duration of response (weeks) 10 26
9 22
9 20
8 19
8 16
60 8 16
8 15
80 7 14
5 13
4 10
B 1·0
0·8
1–Probability of an event
Log-rank p=0·1951 0·6
0·4
0·2
0 0 Number at risk Placebo Rituximab
20
40
60
80
Duration of complete response (weeks) 21 28
19 28
16 28
15 26
14 24
13 22
12 19
10 19
10 17
8 16
8 15
7 15
5 14
4 11
Figure 5: Time to relapse in patients who achieved initial response (A) and patients who achieved complete response (B) Relapse was defined as platelet counts lower than 30 × 10⁹ platelets per L after achieving a response.
between week 25 and the end of the study were 156 g and 167 g, respectively. Total corticosteroid dose did not differ in the two study groups (p=0·33). More than half of patients received no corticosteroids after week 6 from randomisation. The appendix shows corticosteroid consumption in responding patients in both groups. Eight patients received intravenous immunoglobulin in the rituximab group and ten did so in the placebo group. 6
Discussion To our knowledge, this trial is the first randomised placebocontrolled study to assess long-term efficacy of rituximab as a second-line treatment for immune thrombocytopenia in adults (panel). The study shows that provision of rituximab to patients not achieving a sustained response to corticosteroids does not significantly reduce the rate of treatment failure compared with placebo.
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
Articles
Rituximab
Placebo
At baseline Data available
49 (100%)
Rituximab (n=55)
Placebo (n=54)
Grades ≤2
Grade 3
Grades ≤2
Grade 3
50 (100%)
Anaemia
2 (4%)
0
3 (6%)
0
Pyrexia
4 (7%)
0
2 (4%)
0
IgG
9·5 (8·1–12·2)
10·0 (7·9–13·6)
IgM
1·2 (0·8–1·8)
1·1 (0·7–1·6)
Influenza
8 (15%)
0
4 (7%)
0
IgA
1·9 (1·4–2·5)
1·8 (1·3–2·8)
Bronchitis
4 (7%)
0
2 (4%)
0
Upper respiratory tract infection
3 (5%)
0
2 (4%)
0
30 (60%)
Headache
5 (9%)
0
2 (4%)
0
12·6 (9·7–13·9)
Throat irritation
8 (5%)
0
1 (2%)
0
At 24 weeks Data available
37 (76%)
IgG
9·4 (8·1–12·1)
IgM
0·89 (0·5–1·3)
1·07 (0·8–1·67)
Cough
1 (2%)
0
3 (6%)
0
IgA
1·8 (0·96–2·6)
1·9 (1·0–3·0)
Rash
3 (5%)
0
2 (4%)
0
Abdominal pain
2 (4%)
0
1 (2%)
2 (4%)
35 (70%)
Pneumonia
0
1 (2%)
0
1 (2%)
11·3 (9·2–13·2)
Appendicitis
0
1 (2%)
0
0
Last available results Data available
45 (92%)
IgG
9·7 (8·1–12·1)
IgM
0·9 (0·5–1·5)
1·1 (0·9–1·6)
Back pain
0
1 (2%)
1 (2%)
1 (2%)
IgA
1·9 (1·0–7·1)
1·8 (0·4–5·3)
Ovarian cyst
0
0
0
1 (2%)
Pelvic pain
0
0
0
1 (2%)
Data are n (%) or median (IQR).
Table 3: Serum concentrations of immunoglobulin in g/L at baseline at week 24 after treatment administration and at end of the study
The aim of the study was to establish whether the use of rituximab would ultimately render splenectomy superfluous, a hard and a clinically relevant endpoint that is the main rationale for use of rituximab in immune thrombocytopenia.9 The study was powered to show a 30% reduction in splenectomy rate—an effect magnitude we assumed would justify the use of a potentially toxic and expensive medication as an alternative to splenectomy. Although no consensus about absolute criteria for splenectomy exists, the predefined criteria used in this study were based on clinical practice at the time this study was initiated. However, the primary endpoint was amended to treatment failure or splenectomy. Four randomised controlled trials have been done for rituximab in immune thrombocytopenia.17–20 One was a small pilot study19 that did not show a significant difference in the rate of treatment failure between rituximab and placebo. Two studies17,18 compared the efficacy of open-label rituximab with dexamethasone to dexamethasone alone assessing response rates in patients with previously untreated immune thrombocytopenia. The combination resulted in significantly higher rates of sustained response at 6 months. The fourth was a small study that compared low-dose rituximab with dexamethasone to dexamethasone alone, mostly in previously untreated immune thrombocytopenia, showing higher response rates with the combination therapy.20 Compared with three of these studies, our trial represents the widely used clinical practice of use of rituximab when corticosteroids have not worked. Because some patients respond to corticosteroids alone, this practice is preferred to upfront combination therapy.4,5,21,22 Our study did not
Data are n (%). No grade 4 adverse events were reported.
Table 4: Adverse events, except bleeding, occurring in more than 5% of the patients or more than grade 2 in severity in any of the treatment groups
show a significant difference in overall and complete responses with the use of rituximab alone. Notably, different criteria for response were used in these studies. Although the rituximab–dexamethasone studies compared response (defined as platelet count >50 × 10⁹ platelets per L) at 24 weeks, our study compared all responses achieved during 78 weeks to accommodate the possibility of delayed responses. Rituximab causes rapid depletion of B-lymphocytes lasting for approximately 6–12 months. Therefore, responses to rituximab are usually reported within 6 months of administration.14,17 This response profile was also the case in our study. At 24 weeks, 50% of the patients treated with rituximab achieved complete response compared with 35% on placebo. These rates are comparable with those reported in the Italian study17 (53% with dexamethasone–rituximab vs 33% for dexamethasone alone, with an achieved platelet count of ≥100 × 10⁹ platelets per L). Although almost all responses in the rituximab group occurred within 24 weeks, responses in the placebo group continued to occur beyond that time, reducing the net benefit to 8% at 78 weeks (figure 2). Delayed responses in the placebo group might be the result of prolonged exposure to corticosteroids or might suggest spontaneous remissions.5 Notably, three spontaneous remissions were encountered after 1 year from randomisation, two in patients treated with rituximab and one in a patient treated with placebo. All three patients were initially classified as not responding to treatment, but were not splenectomised and did not receive any disease-modulating treatment for months before normalisation of platelet counts. More than half of the responding patients lost their responses during the study. This is in line with other
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
7
Articles
Panel: Research in context Systematic review At the time of writing the study protocol in 2005, we searched Medline for trials and systematic reviews about rituximab in immune thrombocytopenic purpura. The following search terms were used: “immune thrombocytopenic purpura”, “rituximab”, “rituxan”, and “mabthera”. Neither systemic reviews nor randomised controlled trials were identified; only single group studies were identified. A new search of the Medline and Embase databases was carried out in June, 2014, to identify randomised controlled trials in immune thrombocytopenic purpura, and four randomised controlled trials were identified.17–20 In three studies, rituximab plus dexamethasone yielded higher response rates at 6 months compared with dexamethasone alone.17,18,20 The fourth was a small pilot study that did not show a significant difference in the rate of treatment failure between rituximab and placebo.19 Interpretation To our knowledge, our study is the first double-masked placebo-controlled trial to assess the long-term efficacy (78 weeks) of rituximab in patients with immune thrombocytopenic purpura who had not achieved a sustained response to corticosteroids. Our study shows that rituximab as second-line treatment for immune thrombocytopenic purpura does not significantly reduce the rate of long-term treatment failure. However, a small effect of rituximab cannot be ruled out, as suggested by a longer duration of response and numerically higher response rates with rituximab than with placebo. Our findings support the present guidelines, which recommend splenectomy as a standard treatment for patients who do not achieve an adequate response to corticosteroids. However, in the era of individualised therapies, rituximab might be an option early in the course of disease in patients who are intolerant or unresponsive to corticosteroids, or need a high corticosteroid dose, because it is recommended to defer splenectomy to 1 year.
studies showing that most of the responding patients relapse within 2 years of rituximab treatment, yielding a sustained response rate of approximately 20–30%.13 Time to relapse in patients who achieved overall response but not complete response was significantly longer in rituximab-treated patients. Generally, relapses occurred earlier in the placebo group, possibly because many of these responses represent an initial response to a high corticosteroid dose but relapsed during dose tapering. Although patients in the placebo group received a higher total corticosteroid dose during first 24 weeks, the total corticosteroid dose did not differ significantly between the two groups during the study. Numerically more bleeding episodes occurred in the placebo group than the rituximab group, but more bleeding episodes were graded as severe with rituximab than with placebo. Although, the difference in time to first infection was not significant, numerically more infections, particularly respiratory tract infections, occurred in the rituximab group (table 2). The two groups overlapped until week 40; however, more infections occurred in the rituximab group during the last 30 weeks (appendix). The response rates, duration of responses, platelet counts, and number of bleeding events were all in favour of rituximab. The benefit of rituximab was more pronounced during the first 30–40 weeks, as shown by earlier responses and longer time to first bleeding event and also less corticosteroid consumption. However, because of continuing relapses and some delayed responses in the placebo 8
group, the long-term benefit of rituximab was substantially reduced. In view of that fact that 13 (24%) of 55 patients receiving rituximab and eight (14%) of 54 patients receiving placebo had sustained responses, the long-term net rate of sustained responses (>78 weeks) provided by rituximab would be about 10%. This study was not powered to test the statistical significance of a 10% difference in effect. Because several parameters suggest rituximab led to favourable outcomes compared with placebo, we believe that this difference is real and not caused by chance. The question is whether this less than anticipated effect is clinically important and justifies the use of rituximab. According to our results, the number needed to treat to achieve a lasting response in one patient would be ten. Moreover, the relapse curves reach no plateau, and we cannot exclude that these curves might cross at a later timepoint. Therefore, we do not believe that our data can support a general recommendation for use of rituximab monotherapy in patients with immune thrombocytopenia who did not respond to corticosteroids. Nevertheless, in the era of individualised therapies, rituximab might be useful for some patients. Splenectomy will remain the standard second-line treatment for these patients because it provides a sustained response in 60–70% of patients,7 but because it is recommended to defer splenectomy to 1 year,9 rituximab might be an option early in the course of the disease in patients who are intolerant to, unresponsive to, or persistently in need of a high corticosteroid dose. Although a combination of rituximab with dexamethasone as upfront treatment or in previously treated patients seems to improve the short-term response, whether combination therapy improves the long-term effect remains unknown.17,18,23 Thrombopoietin receptor agonists are an alternative to rituximab or splenectomy. Their limitations include the need for continuous treatment to achieve response and the high cost. Furthermore, these drugs are reimbursed in Europe only for patients with chronic immune thrombocytopenia purpura who did not repond to splenectomy, relapse after initial response, or are unfit for the procedure.9 Baseline characteristics were fairly balanced between the two groups apart from in the duration of immune thrombocytopenia before randomisation, which was longer in the placebo group, and for the number of patients actively receiving corticosteroids, which was higher in the rituximab group. The duration of immune thrombocytopenia might have affected the treatment outcome in favour of rituximab. Our change of the primary endpoint is a limitation. Although treatment failure is a less stringent endpoint than splenectomy, nonetheless it accounts for individuals who refused or had a contraindication to splenectomy and needed another treatment. Another limitation was that patients were included at different stages of the disease, with almost half of the patients included having chronic immune thrombocytopenia. Shorter duration of disease has been shown to be associated with an improved
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
Articles
response to rituximab.14,24 Because this is a randomised trial, we do not believe inclusion of patients at different stages has affected our results. Finally, treatment with corticosteroids was not standardised. This factor prevents us from making definite conclusions regarding the exact contribution of rituximab. The strengths of this study include its representative study population, the long follow-up, which enabled the study to capture delayed responses and relapses, and the placebo-controlled design. This study emphasises once again that the implementation of new therapies should not be based on uncontrolled studies, and that every effort should be made to do randomised controlled trials to assess the efficacy and safety of treatments before use. This placebo-controlled study showed that rituximab as a second-line treatment for immune thrombocytopenia did not significantly reduce the rate of long-term treatment failure. However, a small but potentially useful effect of rituximab cannot be ruled out, as indicated by the apparently longer duration of response and higher response rates with rituximab. In the future, combination therapies using several immune suppressing agents should be explored to establish whether such approaches could provide more durable responses. Contributors WG conceived, designed and coordinated the study, recruited patients, participated in data analysis and wrote the study report. PAH designed and coordinated the study, recruited patients, and participated in writing the final report. AW, MM, NBR, and AK participated in screening and recruiting patients into the study, did clinical assessments during follow-up, and were responsible for data collection. GET, JK, and AW participated in study design and follow-up and helped to write the report. BD participated in study design and was responsible for statistical analysis. All authors have seen and approved the final version of the manuscript. RITP study group Monia Khanfir, Sigbjørn Berentsen, Erik Blickfeldt, Peter Meyer, Elin Osvik Velle, Frode Ramslien, Per Morten Sandset, Fredrik Schjesvold, Vigdis Steenberg, Jon Hjalmar Sørbø, Jon Magnus Tangen, Anders Vik, Finn Wisløff, and Moez Elloumi. Declaration of interests WG reports grants from South-East Regional Health Authority, Norway during the conduct of the study; grants, personal fees and non-financial support from Roche; grants and personal fees from Amgen; and personal fees from GlaxoSmithKline. PAH reports grants from South-East Regional Health Authority. MM reports grants and personal fees from Roche and Novartis and personal fees from Amgen, GlaxoSmithKline, Alexion, and Novartis. AK reports personal fees from GlaxoSmithKline. AW, GET, NBR, JK, and BD declare no competing interests. Acknowledgments This trial was funded by South-East Regional Health Authority and Østfold Hospital, Norway. Roche, France provided rituximab free of charge to patients recruited in France and Tunisia. Roche, Norway partially reimbursed rituximab in Norway. References 1 Cines DB, Bussel JB, Liebman HA, Luning Prak ET. The ITP syndrome: pathogenic and clinical diversity. Blood 2009; 113: 6511–21. 2 Provan D, Stasi R, Newland AC, et al. International consensus report on the investigation and management of primary immune thrombocytopenia. Blood 2010; 115: 168–86. 3 Neunert C, Lim W, Crowther M, and the American Society of Hematology. The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood 2011; 117: 4190–207.
4
5
6
7
8
9
10
11 12
13
14
15
16
17
18
19
20
21
22
23
24
Cheng Y, Wong RS, Soo YO, et al. Initial treatment of immune thrombocytopenic purpura with high-dose dexamethasone. N Engl J Med 2003; 349: 831–36. Sailer T, Lechner K, Panzer S, Kyrle PA, Pabinger I. The course of severe autoimmune thrombocytopenia in patients not undergoing splenectomy. Haematologica 2006; 91: 1041–45. Portielje JE, Westendorp RG, Kluin-Nelemans HC, Brand A. Morbidity and mortality in adults with idiopathic thrombocytopenic purpura. Blood 2001; 97: 2549–54. Kojouri K, Vesely SK, Terrell DR, George JN. Splenectomy for adult patients with idiopathic thrombocytopenic purpura: a systematic review to assess long-term platelet count responses, prediction of response, and surgical complications. Blood 2004; 104: 2623–34. Kristinsson SY, Gridley G, Hoover RN, Check D, Landgren O. Long-term risks after splenectomy among 8 149 cancer-free American veterans: a cohort study with up to 27 years follow-up. Haematologica 2014; 99: 392–98. Ghanima W, Godeau B, Cines DB, Bussel JB. How I treat immune thrombocytopenia: the choice between splenectomy or a medical therapy as a second-line treatment. Blood 2012; 120: 960–69. George JN, Woolf SH, Raskob GE, et al. Idiopathic thrombocytopenic purpura: a practice guideline developed by explicit methods for the American Society of Hematology. Blood 1996; 88: 3–40. Rodeghiero F, Ruggeri M. Is splenectomy still the gold standard for the treatment of chronic ITP? Am J Hematol 2008; 83: 91. Arnold DM, Dentali F, Crowther MA, et al. Systematic review: efficacy and safety of rituximab for adults with idiopathic thrombocytopenic purpura. Ann Intern Med 2007; 146: 25–33. Patel VL, Mahévas M, Lee SY, et al. Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia. Blood 2012; 119: 5989–95. Cooper N, Stasi R, Cunningham-Rundles S, et al. The efficacy and safety of B-cell depletion with anti-CD20 monoclonal antibody in adults with chronic immune thrombocytopenic purpura. Br J Haematol 2004; 125: 232–39. Health UNCIo. Common Terminology Criteria for Adverse Events v3.0 2006. http://ctep.cancer.gov/protocoldevelopment/electronic_ applications/docs/ctcaev3.pdf2006 (accessed Feb 20, 2014). Khellaf M, Michel M, Schaeffer A, Bierling P, Godeau B. Assessment of a therapeutic strategy for adults with severe autoimmune thrombocytopenic purpura based on a bleeding score rather than platelet count. Haematologica 2005; 90: 829–32. Zaja F, Baccarani M, Mazza P, et al. Dexamethasone plus rituximab yields higher sustained response rates than dexamethasone monotherapy in adults with primary immune thrombocytopenia. Blood 2010; 115: 2755–62. Gudbrandsdottir S, Birgens HS, Frederiksen H, et al. Rituximab and dexamethasone vs dexamethasone monotherapy in newly diagnosed patients with primary immune thrombocytopenia. Blood 2013; 121: 1976–81. Arnold DM, Heddle NM, Carruthers J, et al. A pilot randomized trial of adjuvant rituximab or placebo for nonsplenectomized patients with immune thrombocytopenia. Blood 2012; 119: 1356–62. Li Z, Mou W, Lu G, et al. Low-dose rituximab combined with short-term glucocorticoids up-regulates Treg cell levels in patients with immune thrombocytopenia. Int J Hematol 2011; 93: 91–98. Neylon AJ, Saunders PW, Howard MR, Proctor SJ, Taylor PR, and the Northern Region Haematology Group. Clinically significant newly presenting autoimmune thrombocytopenic purpura in adults: a prospective study of a population-based cohort of 245 patients. Br J Haematol 2003; 122: 966–74. Mazzucconi MG, Fazi P, Bernasconi S, et al, and the Gruppo Italiano Malattie EMatologiche dell’Adulto (GIMEMA) Thrombocytopenia Working Party. Therapy with high-dose dexamethasone (HD-DXM) in previously untreated patients affected by idiopathic thrombocytopenic purpura: a GIMEMA experience. Blood 2007; 109: 1401–07. Bussel JB, Lee CS, Seery C, et al. Rituximab and 3 dexamethasone cycles provide responses similar to splenectomy in women and those with immune thrombocytopenia less than 2 years duration. Haematologica 2014; 99: 1264–71. Zaja F, Vianelli N, Battista M, et al. Earlier administration of rituximab allows higher rate of long-lasting response in adult patients with autoimmune thrombocytopenia. Exp Hematol 2006; 34: 571–72.
www.thelancet.com Published online February 5, 2015 http://dx.doi.org/10.1016/S0140-6736(14)61495-1
9