CLINICAL PHARMACOLOGY
Mechanism Of Action
Rituximab-pvvr is a monoclonal antibody. Rituximab
products target the CD20 antigen expressed on the surface of pre-B and mature
B-lymphocytes. Upon binding to CD20, rituximab products mediate B-cell lysis.
Possible mechanisms of cell lysis include complement dependent cytotoxicity
(CDC) and antibody dependent cell mediated cytotoxicity (ADCC).
Pharmacodynamics
Non-Hodgkin's Lymphoma (NHL)
In NHL patients, administration of rituximab resulted in
depletion of circulating and tissue-based B cells. Among 166 patients in NHL
Study 1 (NCT000168740), circulating CD19-positive B cells were depleted within
the first three weeks with sustained depletion for up to 6 to 9 months post
treatment in 83% of patients. B-cell recovery began at approximately 6 months
and median B-cell levels returned to normal by 12 months following completion
of treatment.
There were sustained and statistically significant
reductions in both IgM and IgG serum levels observed from 5 through 11 months
following rituximab administration; 14% of patients had IgM and/or IgG serum
levels below the normal range.
Granulomatosis With Polyangiitis (GPA) (Wegener's
Granulomatosis) And Microscopic Polyangiitis
In GPA and MPA patients in GPA/MPA Study 1, peripheral
blood CD19 B-cells depleted to less than 10 cells/μL following the first
two infusions of rituximab, and remained at that level in most (84%) patients through
Month 6. By Month 12, the majority of patients (81%) showed signs of B-cell
return with counts >10 cells/μL. By Month 18, most patients (87%) had
counts >10 cells/μL.
In GPA/MPA Study 2 where patients received
non-U.S.-licensed rituximab as two 500 mg intravenous infusions separated by
two weeks, followed by a 500 mg intravenous infusion at Month 6, 12, and 18,
70% (30 out of 43) of the rituximab-treated patients with CD19+ peripheral B
cells evaluated post-baseline had undetectable CD19+ peripheral B cells at
Month 24. At Month 24, all 37 patients with evaluable baseline CD19+ peripheral
B cells and Month 24 measurements had lower CD19+ B cells relative to baseline.
Pharmacokinetics
Non-Hodgkin's Lymphoma (NHL)
Pharmacokinetics were characterized in 203 NHL patients
receiving 375 mg/m² rituximab weekly by intravenous infusion for 4 doses.
Rituximab was detectable in the serum of patients 3 to 6 months after
completion of treatment.
The pharmacokinetic profile of rituximab when
administered as 6 infusions of 375 mg/m² in combination with 6 cycles of CHOP
chemotherapy was similar to that seen with rituximab alone.
Based on a population pharmacokinetic analysis of data
from 298 NHL patients who received rituximab once weekly or once every three
weeks, the estimated median terminal elimination half-life was 22 days (range,
6.1 to 52 days). Patients with higher CD19-positive cell counts or larger
measurable tumor lesions at pretreatment had a higher clearance. However, dose
adjustment for pretreatment CD19 count or size of tumor lesion is not
necessary. Age and gender had no effect on the pharmacokinetics of rituximab.
Pharmacokinetics were characterized in 21 patients with
CLL receiving rituximab according to the recommended dose and schedule. The
estimated median terminal half-life of rituximab was 32 days (range, 14 to 62
days).
Granulomatosis With Polyangiitis (GPA) (Wegener's
Granulomatosis) And Microscopic Polyangiitis
Based on the population pharmacokinetic analysis of data
in 97 GPA and MPA patients in the GPA/MPA Study 1, who received 375 mg/m² rituximab
once weekly by intravenous infusion for four weeks, the estimated median
terminal elimination half-life was 23 days (range, 9 to 49 days). Rituximab
mean clearance and volume of distribution were 0. 312 L/day (range, 0.115 to
0.728 L/day) and 4.50 L (range, 2.21 to 7.52 L) respectively. Male patients and
patients with higher BSA or positive anti-rituximab antibody levels have higher
clearance. However, further dose adjustment based on gender or anti-rituximab
antibody status is not necessary.
Specific Populations
The pharmacokinetics of rituximab products have not been
studied in children and adolescents.
No formal studies were conducted to examine the effects
of either renal or hepatic impairment on the pharmacokinetics of rituximab
products.
Drug Interaction Studies
Formal drug interaction studies have not been performed
with rituximab products.
Clinical Studies
Relapsed Or Refractory, Low-Grade Or Follicular,
CD20-Positive, B-Cell NHL
The safety and effectiveness of rituximab in relapsed,
refractory CD20+ NHL were demonstrated in 3 single-arm studies enrolling 296
patients.
NHL Study 1
A multicenter, open-label, single-arm study was conducted
in 166 patients with relapsed or refractory, low-grade or follicular, B-cell
NHL who received 375 mg/m² of rituximab given as an intravenous infusion weekly
for 4 doses. Patients with tumor masses >10 cm or with >5,000
lymphocytes/μL in the peripheral blood were excluded from the study.
Results are summarized in Table 3. The median time to
onset of response was 50 days. Disease-related signs and symptoms (including
B-symptoms) resolved in 64% (25/39) of those patients with such symptoms at
study entry.
NHL Study 2
In a multicenter, single-arm study, 37 patients with
relapsed or refractory, low-grade NHL received 375 mg/m² of rituximab weekly
for 8 doses. Results are summarized in Table 3.
NHL Study 3
In a multicenter, single-arm study, 60 patients received
375 mg/m² of rituximab weekly for 4 doses. All patients had relapsed or
refractory, low-grade or follicular, B-cell NHL and had achieved an objective
clinical response to rituximab administered 3.8-35.6 months (median 14.5
months) prior to retreatment with rituximab. Of these 60 patients, 5 received
more than one additional course of rituximab. Results are summarized in Table
3.
Bulky Disease
In pooled data from studies 1 and 3, 39 patients with
bulky (single lesion >10 cm in diameter) and relapsed or refractory,
low-grade NHL received rituximab 375 mg/m² weekly for 4 doses. Results are
summarized in Table 3.
Table 3 : Summary of Rituximab NHL Efficacy Data by
Schedule and Clinical Setting
|
NHL Study 1 Weekly x 4
N=166 |
NHL Study 2 Weekly x 8
N=37 |
NHL Study 1 and NHL Study 3 Bulky disease, Weekly x 4
N=39a |
NHL Study 3 Retreatment, Weekly x 4
N=60 |
Overall Response Rate |
48% |
57% |
36% |
38% |
Complete Response Rate |
6% |
14% |
3% |
10% |
Median Duration of Responseb,c,d |
11.2 |
13.4 |
6.9 |
15.0 |
(Months) [Range] |
[1.9 to 42.1+] |
[2.5 to 36.5+] |
[2.8 to 25.0+] |
[3.0 to 25.1+] |
a Six of these patients are included in the
first column. Thus, data from 296 intent-to-treat patients are provided in this
table.
b Kaplan-Meier projected with observed range.
c“+” indicates an ongoing response.
d Duration of response: interval from the onset of response to
disease progression. |
Previously Untreated, Low-Grade Or Follicular,
CD20-Positive, B-Cell NHL
The safety and effectiveness of rituximab in previously
untreated, low-grade or follicular, CD20+ NHL were demonstrated in 3
randomized, controlled trials enrolling 1,662 patients.
NHL Study 4
A total of 322 patients with previously untreated
follicular NHL were randomized (1:1) to receive up to eight 3-week cycles of
CVP chemotherapy alone (CVP) or in combination with rituximab 375 mg/m² on Day
1 of each cycle (R-CVP) in an open-label, multicenter study. The main outcome
measure of the study was progression-free survival (PFS) defined as the time
from randomization to the first of progression, relapse, or death.
Twenty-six percent of the study population was >60
years of age, 99% had Stage III or IV disease, and 50% had an International
Prognostic Index (IPI) score ≥2. The results for PFS as determined by a
blinded, independent assessment of progression are presented in Table 4. The
point estimates may be influenced by the presence of informative censoring. The
PFS results based on investigator assessment of progression were similar to
those obtained by the independent review assessment.
Table 4 : Efficacy Results in NHL Study 4
|
Study Arm |
R-CVP
N=162 |
CVP
N=160 |
Median PFS (years)a |
2.4 |
1.4 |
Hazard ratio (95% CI)b |
0.44 (0.29, 0.65) |
a p <0.0001, two-sided stratified log-rank
test.
b Estimates of Cox regression stratified by center. |
An open-label, multicenter, randomized (1:1) study was
conducted in 1,018 patients with previously untreated follicular NHL who
achieved a response (CR or PR) to rituximab in combination with chemotherapy.
Patients were randomized to rituximab as single-agent maintenance therapy, 375
mg/m² every 8 weeks for up to 12 doses or to observation. Rituximab was
initiated at 8 weeks following completion of chemotherapy. The main outcome measure
of the study was progression-free survival (PFS), defined as the time from
randomization in the maintenance/observation phase to progression, relapse, or
death, as determined by independent review.
Of the randomized patients, 40% were ≥60 years of age,
70% had Stage IV disease, 96% had ECOG performance status (PS) 0-1, and 42% had
FLIPI scores of 3-5. Prior to randomization to maintenance therapy, patients
had received R-CHOP (75%), R-CVP (22%), or R-FCM (3%); 71% had a complete or
unconfirmed complete response and 28% had a partial response.
PFS was longer in patients randomized to rituximab as
single agent maintenance therapy (HR: 0.54, 95% CI: 0.42, 0.70). The PFS
results based on investigator assessment of progression were similar to those
obtained by the independent review assessment.
Figure 1 : Kaplan-Meier Plot of IRC Assessed PFS
NHL Study 6
A total of 322 patients with previously untreated
low-grade, B-cell NHL who did not progress after 6 or 8 cycles of CVP
chemotherapy were enrolled in an open-label, multicenter, randomized trial.
Patients were randomized (1:1) to receive rituximab, 375 mg/m² intravenous
infusion, once weekly for 4 doses every 6 months for up to 16 doses or no
further therapeutic intervention. The main outcome measure of the study was
progression-free survival defined as the time from randomization to
progression, relapse, or death. Thirty-seven percent of the study population
was >60 years of age, 99% had Stage III or IV disease, and 63% had an IPI
score ≥2.
There was a reduction in the risk of progression,
relapse, or death (hazard ratio estimate in the range of 0.36 to 0.49) for
patients randomized to rituximab as compared to those who received no
additional treatment.
Diffuse Large B-Cell NHL (DLBCL)
The safety and effectiveness of rituximab were evaluated
in three randomized, active-controlled, open-label, multicenter studies with a
collective enrollment of 1,854 patients. Patients with previously untreated
diffuse large B-cell NHL received rituximab in combination with
cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) or other
anthracycline-based chemotherapy regimens.
NHL Study 7
A total of 632 patients age ≥60 years with DLBCL
(including primary mediastinal B-cell lymphoma) were randomized in a 1:1 ratio
to treatment with CHOP or R-CHOP. Patients received 6 or 8 cycles of CHOP, each
cycle lasting 21 days. All patients in the R-CHOP arm received 4 doses of rituximab
375 mg/m² on Days -7 and -3 (prior to Cycle 1) and 48-72
hours prior to Cycles 3 and 5. Patients who received 8 cycles of CHOP also
received rituximab prior to Cycle 7. The main outcome measure of the study was
progression-free survival, defined as the time from randomization to the first
of progression, relapse, or death. Responding patients underwent a second
randomization to receive rituximab or no further therapy.
Among all enrolled patients, 62% had centrally confirmed
DLBCL histology, 73% had Stage III-IV disease, 56% had IPI scores
≥2, 86% had ECOG performance status of <2, 57% had elevated LDH
levels, and 30% had two or more extranodal disease sites involved. Efficacy
results are presented in Table 5. These results reflect a statistical approach
which allows for an evaluation of rituximab administered in the induction
setting that excludes any potential impact of rituximab given after the second
randomization.
Analysis of results after the second randomization in NHL
Study 7 demonstrates that for patients randomized to R-CHOP, additional
rituximab exposure beyond induction was not associated with further improvements
in progression-free survival or overall survival.
NHL Study 8
A total of 399 patients with DLBCL, age ≥60 years,
were randomized in a 1:1 ratio to receive CHOP or R-CHOP. All patients received
up to eight 3-week cycles of CHOP induction; patients in the R-CHOP arm
received rituximab 375 mg/m² on Day 1 of each cycle. The main outcome measure
of the study was event-free survival, defined as the time from randomization to
relapse, progression, change in therapy, or death from any cause. Among all
enrolled patients, 80% had Stage III or IV disease, 60% of patients had an
age-adjusted IPI ≥2, 80% had ECOG performance status scores <2, 66%
had elevated LDH levels, and 52% had extranodal involvement in at least two
sites. Efficacy results are presented in Table 5.
NHL Study 9
A total of 823 patients with DLBCL, aged 18-60
years, were randomized in a 1:1 ratio to receive an anthracycline-containing
chemotherapy regimen alone or in combination with rituximab. The main outcome
measure of the study was time to treatment failure, defined as time from
randomization to the earliest of progressive disease, failure to achieve a
complete response, relapse, or death. Among all enrolled patients, 28% had
Stage III-IV disease, 100% had IPI scores of ≤1, 99% had ECOG
performance status of <2, 29% had elevated LDH levels, 49% had bulky
disease, and 34% had extranodal involvement. Efficacy results are presented in
Table 5.
Table 5 : Efficacy Results in NHL Studies 7, 8, and 9
Main outcome |
NHL Study 7
(n=632) |
NHL Study 8
(n=399) |
NHL Study 9
(n=823) |
R-CHOP |
CHOP |
R-CHOP |
CHOP |
R-Chemo |
Chemo |
Progression-free survival (years) |
Event-free survival (years) |
Time to treatment failure (years) |
Median of main outcome measure |
3.1 |
1.6 |
2.9 |
1.1 |
NEb |
NEb |
Hazard ratiod |
0.69a |
0.60a |
0.45a |
Overall survival at 2 yearsc |
74% |
63% |
69% |
58% |
95% |
86% |
Hazard ratiod |
0.72a |
0.68a |
0.40a |
a Significant at p <0.05, 2-sided.
b NE=Not reliably estimable.
cKaplan-Meier estimates.
d R-CHOP vs. CHOP. |
In NHL Study 8, overall survival estimates at 5 years
were 58% vs. 46% for R-CHOP and CHOP, respectively.
Ninety-Minute Infusions In Previously Untreated Follicular
NHL And DLBCL
In NHL Study 10, a total of 363 patients with previously
untreated follicular NHL (n=113) or DLBCL (n=250) were evaluated in a
prospective, open-label, multi-center, single-arm trial for the safety of
90-minute rituximab infusions. Patients with follicular NHL received rituximab
375 mg/m² plus CVP chemotherapy. Patients with DLBCL received rituximab 375
mg/m² plus CHOP chemotherapy. Patients with clinically significant cardiovascular
disease were excluded from the study. Patients were eligible for a 90-minute
infusion at Cycle 2 if they did not experience a Grade 3-4 infusion-related
adverse event with Cycle 1 and had a circulating lymphocyte count <5,000/mm³
before Cycle 2. All patients were pre-medicated with acetaminophen and an
antihistamine and received the glucocorticoid component of their chemotherapy
prior to rituximab infusion. The main outcome measure was the development of
Grade 3-4 infusion-related reactions on the day of, or day after, the 90-minute
infusion at Cycle 2 [see ADVERSE REACTIONS].
Eligible patients received their Cycle 2 rituximab
infusion over 90 minutes as follows: 20% of the total dose given in the first
30 minutes and the remaining 80% of the total dose given over the next 60
minutes [see DOSAGE AND ADMINISTRATION]. Patients who tolerated the
90-minute rituximab infusion at Cycle 2 continued to receive subsequent
rituximab infusions at the 90-minute infusion rate for the remainder of the
treatment regimen (through Cycle 6 or Cycle 8).
The incidence of Grade 3-4 infusion-related reactions at
Cycle 2 was 1.1% (95% CI [0.3%, 2.8%]) among all patients, 3.5% (95% CI [1.0%,
8.8%]) for those patients treated with R-CVP, and 0.0% (95% CI [0.0%, 1.5%])
for those patients treated with R-CHOP. For Cycles 2-8, the incidence of Grade
3-4 infusion-related reactions was 2.8% (95% CI [1.3%, 5.0%]). No acute fatal
infusion-related reactions were observed.
Chronic Lymphocytic Leukemia (CLL)
The safety and effectiveness of rituximab were evaluated
in two randomized (1:1) multicenter open-label studies comparing FC alone or in
combination with rituximab for up to 6 cycles in patients with previously
untreated CLL [CLL Study 11 (n=817)] or previously treated CLL [CLL Study 12
(n=552)]. Patients received fludarabine 25 mg/m²/day and cyclophosphamide 250
mg/m²/day on days 1, 2, and 3 of each cycle, with or without rituximab. In both
studies, seventy-one percent of CLL patients received 6 cycles and 90% received
at least 3 cycles of rituximab-based therapy.
In CLL Study 11, 30% of patients were 65 years or older,
31% were Binet stage C, 45% had B symptoms, more than 99% had ECOG performance
status (PS) 0-1, 74% were male, and 100% were White. In CLL Study 12,
44% of patients were 65 years or older, 28% had B symptoms, 82% received a
prior alkylating drug, 18% received prior fludarabine, 100% had ECOG PS 0-1,
67% were male, and 98% were White.
The main outcome measure in both studies was
progression-free survival (PFS), defined as the time from randomization to
progression, relapse, or death, as determined by investigators (CLL Study 11)
or an independent review committee (CLL Study 12). The investigator assessed
results in CLL Study 12 were supportive of those obtained by the independent
review committee. Efficacy results are presented in Table 6.
Table 6 : Efficacy Results in CLL Studies 11 and 12
|
CLL Study 11* (Previously untreated) |
CLL Study 12* (Previously treated) |
R-FC
N=408 |
FC
N=409 |
R-FC
N=276 |
FC
N=276 |
Median PFS (months) |
39.8 |
31.5 |
26.7 |
21.7 |
Hazard ratio (95% CI) |
0.56 (0.43,0.71) |
0.76 (0.6,0.96) |
P value (Log-Rank test) |
<0.01 |
0.02 |
Response rate (95% CI) |
86% (82, 89) |
73% (68, 77) |
54% (48, 60) |
45% (37, 51) |
* As defined in 1996 National Cancer Institute Working
Group guidelines. |
Across both studies, 243 of 676 rituximab-treated
patients (36%) were 65 years of age or older and 100 rituximab-treated patients
(15%) were 70 years of age or older. The results of exploratory subset analyses
in elderly patients are presented in Table 7.
Table 7 : Efficacy Results in CLL Studies 11
and 12 in Subgroups Defined by Agea
Age subgroup |
CLL Study 11 |
CLL Study 12 |
Number of Patients |
Hazard Ratio for PFS (95% CI) |
Number of Patients |
Hazard Ratio for PFS (95% CI) |
Age <65 yrs |
572 |
0.52
(0.39, 0.70) |
313 |
0.61
(0.45, 0.84) |
Age ≥65 yrs |
245 |
0.62
(0.39, 0.99) |
233 |
0.99
(0.70, 1.40) |
Age <70 yrs |
736 |
0.51
(0.39, 0.67) |
438 |
0.67
(0.51, 0.87) |
Age ≥70 yrs |
81 |
1.17
(0.51, 2.66) |
108 |
1.22
(0.73, 2.04) |
a From exploratory analyses. |
Granulomatosis With Polyangiitis (GPA) (Wegener's Granulomatosis)
And Microscopic Polyangiitis (MPA)
Induction Treatment of Patients with Active Disease
(GPA/MPA Study 1)
A total of 197 patients with active, severe GPA and MPA
(two forms of ANCA Associated Vasculitides) were treated in a randomized,
double-blind, active-controlled multicenter, non-inferiority study, conducted
in two phases â⬓ a 6 month remission induction phase and a 12 month remission
maintenance phase. Patients were 15 years of age or older, diagnosed with GPA
(75% of patients) or MPA (24% of patients) according to the Chapel Hill
Consensus conference criteria (1% of the patients had unknown vasculitis type).
All patients had active disease, with a Birmingham Vasculitis Activity Score
for Granulomatosis with Polyangiitis (BVAS/GPA) ≥3, and their disease was
severe, with at least one major item on the BVAS/GPA. Ninety-six (49%) of
patients had new disease and 101 (51%) of patients had relapsing disease.
Patients in both arms received 1,000 mg of pulse
intravenous methylprednisolone per day for 1 to 3 days within 14 days prior to
initial infusion. Patients were randomized in a 1:1 ratio to receive either
rituximab 375 mg/m² once weekly for 4 weeks or oral cyclophosphamide 2 mg/kg
daily for 3 to 6 months in the remission induction phase. Patients were
pre-medicated with antihistamine and acetaminophen prior to rituximab infusion.
Following intravenous corticosteroid administration, all patients received oral
prednisone (1 mg/kg/day, not exceeding 80 mg/day) with pre-specified tapering.
Once remission was achieved or at the end of the 6 month remission induction
period, the cyclophosphamide group received azathioprine to maintain remission.
The rituximab group did not receive additional therapy to maintain remission.
The main outcomemeasure for both GPA and MPA patients was achievement of
complete remission at 6 months defined as a BVAS/GPA of 0, and off
glucocorticoid therapy. The pre-specified non-inferiority margin was a
treatment difference of 20%. As shown in Table 8, the study demonstrated non-inferiority
of rituximab to cyclophosphamide for complete remission at 6 months.
Table 8 : Percentage of Patients Who Achieved Complete
Remission at 6 Months (Intent-to-Treat Population)
|
Rituximab
(n=99) |
Cyclophosphamide
(n=98) |
Treatment Difference (Rituximab - Cyclophosphamide) |
Rate |
64% |
53% |
11% |
95.1%b CI |
(54%, 73%) |
(43%, 63%) |
(-3%, 24%)a |
a Non-inferiority was demonstrated because the
lower bound was higher than the prespecified non-inferiority margin (-3%
>-20%).
b The 95.1% confidence level reflects an additional 0.001 alpha to
account for an interim efficacy analysis. |
Complete Remission (CR) At 12 And 18 Months
In the rituximab group, 44% of patients achieved CR at 6
and 12 months, and 38% of patients achieved CR at 6, 12, and 18 months. In
patients treated with cyclophosphamide (followed by azathioprine for
maintenance of CR), 38% of patients achieved CR at 6 and 12 months, and 31% of
patients achieved CR at 6, 12, and 18 months.
Retreatment Of Flares With Rituximab
Based upon investigator judgment, 15 patients received a
second course of rituximab therapy for treatment of relapse of disease activity
which occurred between 8 and 17 months after the induction treatment course of
rituximab.
Follow Up Treatment Of Patients With GPA/MPA Who Have
Achieved Disease Control With Other Immunosuppressant (GPA/MPA Study 2)
A total of 115 patients (86 with GPA, 24 with MPA, and 5
with renal-limited ANCA-associated vasculitis) in disease remission were
randomized to receive azathioprine (58 patients) or non-U.S.-licensed rituximab
(57 patients) in this open-label, prospective, multi-center, randomized,
active-controlled study. Eligible patients were 21 years and older and had
either newly diagnosed (80%) or relapsing disease (20%). A majority of the
patients were ANCA-positive. Remission of active disease was achieved using a
combination of glucocorticoids and cyclophosphamide. Within a maximum of 1
month after the last cyclophosphamide dose, eligible patients (based on BVAS of
0), were randomized in a 1:1 ratio to receive either non-U.S.-licensed
rituximab or azathioprine.
The non-U.S.-licensed rituximab was administered as two
500 mg intravenous infusions separated by two weeks (on Day 1 and Day 15)
followed by a 500 mg intravenous infusion every 6 months for 18 months.
Azathioprine was administered orally at a dose of 2 mg/kg/day for 12 months,
then 1.5 mg/kg/day for 6 months, and finally 1 mg/kg/day for 4 months;
treatment was discontinued after 22 months. Prednisone treatment was tapered
and then kept at a low dose (approximately 5 mg per day) for at least 18 months
after randomization. Prednisone dose tapering and the decision to stop
prednisone treatment after month 18 were left at the investigator's discretion.
Planned follow-up was until month 28 (10 or 6 months,
respectively, after the last non-U.S.-licensed rituximab infusion or
azathioprine dose). The primary endpoint was the occurrence of major relapse
(defined by the reappearance of clinical and/or laboratory signs of vasculitis
activity that could lead to organ failure or damage, or could be life
threatening) through month 28.
By month 28, major relapse occurred in 3 patients (5%) in
the non-U.S.-licensed rituximab group and 17 patients (29%) in the azathioprine
group.
The observed cumulative incidence rate of first major
relapse during the 28 months was lower in patients on non-U.S.-licensed
rituximab relative to azathioprine (Figure 2).
Figure 2 : Cumulative Incidence Over Time of First
Major Relapse