Methods
Study design
EARLY is a prospective, randomized, actively controlled, double-blind and
parallel group multi-centre Phase III clinical trial with a total of 11 German
study sites involved. The trial was classified as Phase III according to the
recommendations of the ethical board and the federal authorities. The reasons
were that this is not a first-in-patient study and was powered for efficacy. The
trial protocol has been approved by the local ethics committees and the medical
regulatory authorities in Germany (Federal Institute for Drugs and Medical
Devices (Bundesinstitut für Arzneimittel und Medizinprodukte (BfArM)).
Prior to the inclusion of the first patient, the study was registered in the EU
Clinical Trials Register (2018-001514-15) and the International Clinical Trials
Registry Platform (ICTRP, http://apps.who). The complete trial
protocol, as submitted to the authorities with free access, is available in the
most recent version at:
https://drks.de/search/de/trial/DRKS00016043
and
https://trialsearch.who.int/Trial2.aspx?TrialID=DRKS00016043.
The Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT)
[21 ]
[22 ] checklist is presented in the Supplement.
Study population
Inclusion criteria were as follows: (1) patients aged between 18 and 65 years,
with (2) signed informed consent and a (3) DSM-V diagnosis of schizophrenia
confirmed by the Mini International Neuropsychiatric Interview [23 ] and (4) at least one documented prior
hospitalization due to the illness in the medical history (the current
hospitalization can be considered as “prior” hospitalization if
its≥4 weeks) at screening. Furthermore, (5) for treatment-naïve
patients (defined as no previous antipsychotic treatment or a maximum of 30 days
of treatment), an antipsychotic treatment attempt of at least 30 days with an
antipsychotic in a therapeutic dose according to local guidelines other than
clozapine and olanzapine before the screening phase is needed. For
non-treatment-naïve patients (defined as having been treated for more
than 30 days with an antipsychotic), discontinuation of a foregoing
antipsychotic treatment prior to the screening phase within a maximum of six
months (=180 days) is possible (corresponding to the estimated average
time for an antipsychotic washout phase and the expected time to develop a
relapse of the disease). For patients being treated with a long-acting
antipsychotic (other than paliperidone palmitate as a 3-monthly injection), an
inclusion is possible if the inclusion date corresponds to the planned date of
the next injection plus five to seven days. For patients being treated with oral
olanzapine, an inclusion is possible if this treatment has lasted for no longer
than 2 weeks prior to inclusion and if exclusion criteria 8 is not fulfilled.
Another inclusion criterion is (6) the clinical need for a medication switch
because of clinical inefficacy or side effects or the clinical need for a
reintroduction of antipsychotic treatment after treatment discontinuation prior
to the screening phase (see 5.). Finally, (7) patients must have moderate
symptomatology on the Positive and Negative Syndrome Scale (PANSS) [24 ], defined as a score≥4 for two
or more symptoms from P1-P7 or a score of≥6 for one symptom from P1-P7
(minimum threshold definition) at screening. Participation in the EARLY trial is
possible for (8) male participants and female participants who are not capable
of bearing children or who use a method of contraception that is medically
approved by the health authority of the respective country at screening.
Exclusion criteria were defined as follows: (1) Patients who are not suitable for
the study in the opinion of the investigator and (2) patients who are unable to
give informed consent as well as (3) coercive treatment at the time of study
inclusion. Furthermore, patients with a (4) white blood cell count (WBC) at
inclusion not meeting the requirements for clozapine use in Germany were
excluded. Patients must have normal leukocyte findings (white blood cell
count≥3500/mm3
(≥3.5×109 /l) and Absolute Neutrophil
Count≥2000/mm3
(≥2.0×109 /l) at the screening visit.
Furthermore, (5) the presence of one or more contraindications against any of
the study drugs as mentioned in the Summary of Product Characteristics (SmPC).
Another exclusion criterion is the presence of (6) treatment-naïve or
treatment-resistant schizophrenia. Treatment-naïve is defined as having
no previous antipsychotic treatment or a maximum of 30 days of treatment.
Treatment-resistance is defined as two antipsychotic trials (with antipsychotics
from two different chemical classes) for a period of≥6 weeks with
chlorpromazine (CPZ) equivalent doses≥600 mg/day, both
of which took place immediately before the screening phase. Furthermore, (7) the
diagnosis of a primary substance dependency other than nicotine is an exclusion
criterion, as well as (8) documented previous non-response to an 8-week drug
trial with olanzapine or any documented previous treatment with clozapine, (9)
intolerance to one of the study drugs and (10) pregnancy (incl. positive blood
pregnancy test)/lactation (female patients).
The aim was to define a study population with a recurrent schizophrenia disease
course not fulfilling the criteria for treatment-resistance. To exclude patients
with treatment-resistance, an operationalized consensus definition was selected
[19 ]. Treatment-naïve patients
were excluded due to the high response and remission rates of these patients to
any antipsychotic treatment.
Intervention
Patients are randomized either to double-blind clozapine or olanzapine and
treated for 8 weeks. Clozapine and olanzapine are, to a certain extent,
comparable in terms of weight gain/appetite increase, motor side
effects, sedation and obstipation – thus, the risk of accidental
unblinding due to compound-specific side effects are reduced as much as
possible, as shown in a head-to-head trial in individuals with
treatment-resistant schizophrenia [25 ].
Moreover, olanzapine is one of the most-effective non-clozapine antipsychotics
[26 ]
[27 ]
[28 ] challenging clozapine
in terms of efficacy in our target population of non-treatment-resistant
schizophrenia patients.
Randomisation
Central randomization will be performed using predefined randomisation lists,
which are stratified by center and created at the Muenchner Studienzentrum (MSZ)
using RANCODE professional (version 3.6, IDV, Gauting, Germany) and permuted
blocks. The randomization lists are forwarded to the pharmacist at the central
pharmacy (University of Heidelberg), who prepares the blinded study medication
accordingly. Both patient and treating physician are blinded during the
double-blind phase.
Patients have the right to withdraw the informed consent at any time with no
reason given and are then declared a drop-out. The investigator, on the other
hand, has the right to exclude a patient from the study in the event of
concomitant disease, adverse events, therapy failure or any other reason or
condition that warrants withdrawal in the interest of the patient. Such patients
are not primarily defined as dropouts. All patients leaving the trial prior to
visit 9 but accepting the assessment of the primary endpoint at the timepoint of
visit 9 (+1 week) are no dropouts.
The study flow-chart is presented in [Fig.
1 ]. The total study period after randomization is 12 weeks for every
patient, including 8 weeks of intervention and 4 weeks of safety follow-up.
Patients finishing the double-blind treatment phase are invited to participate
in the naturalistic extension study that is not part of the primary study (see
[Table 4 ]). To be included in the
naturalistic follow-up after 12 weeks, patients will have to sign an additional
consent form.
Fig. 1 Schematic study flow chart.; BL: baseline; FU: Follow-up;
PANSS RSWG items: remission criteria according to the Remission in
Schizophrenia Working Group; V1-V13: Visit 1 - Visit 13.
Table 4 Naturalistic follow-up visits
N1–N3.
Study Visit
N1
N2
N3
Day
+26 weeks (D 238+/– 4 weeks)
after V9
+52 weeks (D 420+/– 4 weeks)
after V9
+104 weeks (D 784+/– 4 weeks)
after V9
Naturalistic follow-up
Medical and psychiatric history
X
X
X
Year of birth, Sex
X
Living circumstance, working status
X
X
X
Actual treatment
X
X
X
PANSS
X
X
X
PANSS RSWG items
X
X
X
CGI
X
X
X
GAF
X
X
X
PSP
X
X
X
Q-LES-Q-18
X
X
X
SF-12
X
X
X
PANSS: Positive and Negative Syndrome Scale; PANSS RSWG items: Remission
criteria according to the Remission in Schizophrenia Working Group; CGI:
Clinical Global Impression scale; GAF: Global Assessment Scale of
Functioning; Q-LES-Q-18: Abbreviated quality of life enjoyment and
satisfaction questionnaire; SF-12: Short Form Health Survey.
Medication
The dose range of clozapine will be 75–600 mg/day and for
olanzapine the dose range will be 2.5–20 mg/day and
eight dosing stages have been defined for the trial: stage 1 (75 mg
clozapine/2.5 mg olanzapine); stage 2 (150 mg
clozapine/5 mg olanzapine); stage 3 (225 mg
clozapine/7.5 mg olanzapine); stage 4 (300 mg
clozapine/10 mg olanzapine); stage 5 (375 mg
clozapine/12.5 mg olanzapine); stage 6 (450 mg
clozapine/15 mg olanzapine); stage 7 (525 mg
clozapine/17.5 mg olanzapine) and stage 8 (600 mg
clozapine/20 mg olanzapine). The detailed titration scheme is
displayed in [Fig. 1 ] and [Tables 1 ]
[2 ]. Dosage can be adjusted to higher or lower dosages within the
listed ranges if a patient fails to improve or if patients develop relevant
side-effects. The maximum clozapine of 600 mg/d clozapine
corresponds [29 ] to the approved maximum
dose of olanzapine (20 mg) but is below the dosages offered in trials
conducted in treatment-resistant patients, which allowed up to
900 mg/day. However, the mean dose actually reached in the
double-blind treatment-resistant patients trials were, e. g.,
400 mg [25 ], 304 mg [30 ] or 291 mg [31 ]; one meta-analysis analysed trials of
treatment-resistant schizophrenia patients and showed that the mean clozapine
dose was only 392 mg/d [32 ]. Furthermore, the patients in our study are not
treatment-resistant, so we do not expect them to require such high doses as did
the treatment-resistant patients in the earlier studies and the selected maximum
dose was chosen to match as good as possible to the maximum dose of
20 mg olanzapine. Our semi-flexible dosing scheme was chosen to allow
patients and study doctors to find an optimal individual dose and allowing to
develop a double-blind titration scheme with the minimum possible amount of
study drug to be manufactured.
Table 1 Stages and dose ranges of the trial. The rules for
using the lowest dosages (clozapine 75 mg/olanzapine
2.5 mg). V=visit.
Stage 1
75 mg clozapine or 2.5 mg olanzapine,
earliest timepoint: V2
Stage 2
150 mg clozapine or 5 mg olanzapine, earliest
timepoint: V2
Stage 3
225 mg clozapine or 7.5 mg olanzapine,
earliest timepoint: V3
Stage 4
300 mg clozapine or 10 mg olanzapine,
earliest timepoint: V3
Stage 5
375 mg clozapine or 12.5 mg olanzapine,
earliest timepoint: V4
Stage 6
450 mg clozapine or 15 mg olanzapine,
earliest timepoint: V4
Stage 7
525 mg clozapine or 17,5 mg olanzapine,
earliest timepoint: V5
Stage 8
600 mg clozapine or 20 mg olanzapine,
earliest timepoint: V5
Table 2
V1 (day 0–1 )
12.5 mg clozapine or 0 mg
olanzapine
V2 (day 7)
75–150 mg clozapine or
2.5–5 mg olanzapine
V3 (day 14)
150–325 mg clozapine or
5–10 mg olanzapine
V4 (day 21)
150–500 mg clozapine or
5–15 mg olanzapine
V5 (day 28)
150–600 mg clozapine or
5–20 mg olanzapine
V6 (day 35)
150–600 mg clozapine or
5–20 mg olanzapine
V7 (day 42)
150–600 mg clozapine or
5–20 mg olanzapine
V8 (day 49)
150–600 mg clozapine or
5–20 mg olanzapine
V9 (day 56)
150–600 mg clozapine or
5–20 mg olanzapine
The titration schemes of both drugs are the same to maintain the blind (see [Tables 1 ]
[2 ]). Dosages should be increased to a maximum of one capsule per day
of the titration blister (25 mg clozapine/capsule or placebo)
from the day after the start of titration during the first week. The dose has to
be increased by using capsules of the titration blister until stage 1
(75 mg clozapine or 2.5 mg olanzapine/day) has been
reached, which should be preferably 3 to 7 days after day 1. Next, dosage should
be up-titrated until stage 2 (150 mg clozapine or 5 mg
olanzapine/day) has been reached. Stage 2 is the minimal target dosage
in this trial and dosage below stage 2 (namely stage 1) is only permitted due to
tolerability reasons and this down titration must be documented. Dosage can be
adjusted to higher or lower dosages if a patient fails to improve or if patients
develop relevant side-effects. Titration can be slowed or stopped below the
target dose if subjects cannot tolerate the standard titration schedule because
of adverse effects. A daily dosage of<2.5 mg (stage 1)
for>7 consecutive days and a daily dosage over 20 mg/d
for>3 consecutive days both constitute a protocol violation. The visit
corresponding to visit 9 must be scheduled for week 8 where RSWG criteria must
be assessed. At visit 9, participants and responsible study doctors will be
asked to estimate whether the given participant was in the clozapine or
olanzapine treatment arm. Unused patient-specific study medication and used
blisters will be returned at visit 9 to assess adherence. Moreover, blinded
blood levels, as detailed below, will also be used to assess adherence.
Comedication
The use of antipsychotics in addition to the study medication is not permitted
from one day after visit 3 until visit 9 (see [Fig. 1 ] and [Table 3 ]). During
the 2-week titration phase after randomization, the on-demand use of haloperidol
(max. 10 mg/day) is permitted as rescue medication in accordance
with previous trials [25 ]. Pre-study
antipsychotics will be tapered down during the titration phase and from day 15
onwards, only study antipsychotics are permitted. Treatment with antidepressants
is only permitted in cases of clinically relevant depression (Calgary Depression
Scale for Schizophrenia sum score>6 with depressive symptoms for a
minimum of 2 weeks) and only if antidepressants are used that do not have
relevant interactions with any of the study drugs. The introduction of
mood-stabilizers during the trial is not permitted. Pretrial mood-stabilizers
can be continued if the used compounds do not have relevant interactions with
any of the study drugs. All drugs to treat somatic conditions are permitted if
the used substances do not have relevant interactions with any of the study
drugs.
Table 3 Frequency and scope of study visits.
Study Visit
Screening ##
Baseline 1 ###
2
3
4
5
6
7
8
9
10
11
12
13
Day
–14 to 0
0–1
7
14
21
28
35
42
49
56 Primary Endpoint*
63
70
77
84 (EOS)
Intervention of medication
Administration from day 1 to day 56
Informed consent
X ****
Inclusion/exclusion criteria
X
Drug screening
X
AUDIT interview and MINI
X
Fagerström test (Smoker)
X
Demographic data/medical/Psychiatric
history
X
Pre-study medication
X
X
X
X
Physical examination
X
X
X
X
X
X
X
X
X
X
Study laboratory
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Pregnancy test
X
Number of cigarettes
X
X
X
X
X
Interaction
check******
X
X
X
X
X
X
X
X
X
PANSS
***,
#
X
X
X
X
PANSS RSWG items
X
X
X
X
X
CGI and GAF
X
X
X
PSP, ISST, Trail-Making-Test (TMT)
X
X
CDSS
X
X
X
Q-LES-Q-18, SF-12, DAI10, SWN-K, TALD
X
X
Attitude towards clinical trials
X
BARS
X
X
GASS
X
X
X
X
St. Hans Rating Scale (SHRS)
X
X
CCCS
X
X
X
X
Assessment blood pressure and heart rate
X
X
X
X
X
X
X
X
X
Assessment of weight and BMI
X
X
X
EEG
X
X
ECG
X
X
X
X
X
X
Study medication (re-)dispensing
X
X
X
X
X
X
X
X
*
Study medication return
X
Blood draw drug levels **
X
X
Assessment of blinding integrity
X
Discontinuation check
X
X
X
X
X
X
X
X
SAE/AE
X
X*****
Concomitant medication
X
X*****
AUDIT: Alcohol Use Disorders Identification Test; MINI: MINI-Interview
for ICD-10 and DSM-V diagnosis; PANSS: Positive and Negative Syndrome
Scale; PANSS RSWG items: Remission criteria according to the Remission
in Schizophrenia Working Group; CGI: Clinical Global Impression scale;
GAF: Global Assessment Scale of Functioning; PSP: Personal and Social
Performance Scale; ISST: InterSePT scale for suicidal thinking; TMT:
Trail-Making-Test, CDSS: Calgary Depression Rating Scale for
Schizophrenia; Q-LES-Q-18: Abbreviated quality of life enjoyment and
satisfaction questionnaire; SF-12: Short Form Health Survey; DAI-10: The
Drug Attitude Inventory; SWN-K: The subjective well-being under
neuroleptic scale; TALD: Thought and Language Disorder Scale; BARS:
Barnes Akathisia Rating Scale; GASS-C: Glasgow Antipsychotic
Side-effects Scale for Clozapine; SHRS: St. Hans Rating Scale; CCCS:
Cleveland Clinic Constipation Score; BMI: Body-Mass Index; EEG:
electroencephalogram; ECG: electrocardiogram; (S)AE: (Serious) Adverse
Event.
* one additional blister of study medication can be distributed
to the patient if the patient is discharged from inpatient treatment
(lasting for around 4 days)
** Documentation of amount and type of capsules
related to the time point of blood draw for drug level assessment; If a
blood draw is not possible on the planned visits, the drug levels can be
assessed at the next planned visit.
*** During the study visits of the
double-blind treatment phase, psychopathological worsening of the
patient is documented if the members of the study team performing the
current study visit AND/OR the current treating physicians at
the time of the study visit suspect a relevant worsening of psychotic
symptoms. If psychopathological worsening is suspected in this
regard, additional PANSS ratings should be performed in those visits
where a PANSS rating is not planned. Psychopathological
worsening will then be assessed and defined as an increase of at least
20% in the PANSS total score. The measure of the increase in
PANSS total score is calculated according to the following formula:
Δ-PANSS total %=(PANSS total score at current
Visit – PANSS total score at Visit Baseline (visit 1))*
100/(PANSS total V Baseline (visit 1) – 30) [%].
Thereby the deterioration in relation to baseline, i. e., visit
1, will be assessed as a measure for worsening throughout the study.
**** Also, in case of relevant
psychopathological worsening as defined by the above-mentioned
criteria (i. e., an increase in PANSS total score of at
least 20%) or an emerging SAE), a clinician that is not
associated with the study (board-certified psychiatrist/Facharzt
für Psychiatrie und Psychotherapie) has to reconfirm and
document that the patient continues to be capable of giving informed
consent. If not, the patient has to be excluded from the study.
***** Concomitant
medication will be reported in case of a new laboratory-related AE after
visit 9/Early Termination or any SAE between Visit
10–13; after completing visit 9, only laboratory-related AEs and
SAEs will be documented.
******Interaction check
at baseline can be performed with textbooks, tables or commercial
software. Interaction checks from visit 2 - visit 9 have not to be
performed with a specific instrument.
# NOTE : PANSS rating scores obtained at the Screening Visit (visit
0, V0) can be used for the Baseline Visit (V1, visit 1) within a maximum
range of 7 days.
## NOTE : If during the course of treatment of inpatients OR
outpatients, an ECG and/or an EEG and/or laboratory
tests have been performed prior to inclusion into the study
(i. e., prior to giving informed consent on the informed consent
(IC) form), these can also be used for the study visit V0 (Screening
Visit) within the following day ranges: An ECG performed within 7 days
before the Screening Visit, can be used for the Screening Visit. EEG
recordings are valid for the Screening Visit if they are performed
within 5 days prior to the day on which the Screening Visit is
performed. Similarly, laboratory tests are valid for the Screening Visit
if they are performed within 3 days prior to the day on which the
Screening Visit is performed.
### NOTE : If an ECG and/or an EEG and/or
laboratory tests have been performed prior to inclusion into the study
(i. e., prior to giving informed consent on the IC form), these
can also be used for visit 1 (Baseline Visit) within the following day
ranges: An ECG performed within 7 days before the Baseline Visit, can be
used both for the Screening and the Baseline Visit. EEG recordings are
valid for the Baseline Visit if they are performed within 5 days prior
to the day on which the Baseline Visit is performed. Similarly,
laboratory tests are valid for the Baseline Visit if they are performed
within 3 days prior to the day on which the Baseline Visit is performed.
Finally, the PANSS rating scores obtained at the Screening Visit can be
used for the Baseline Visit within a maximum range of 7 days.
Blood levels
Blinded blood levels of clozapine (and desmethylclozapine) and olanzapine
(desmethylolanzapine) will be measured 2 and 4 weeks after randomization at
visits 3 and 5 at the central laboratory of the Klinikum der
Universität München. If the assessment of the blood levels is
not possible at these visits, the respective blood draw for blood level analyses
should be performed at the next planned visit. All sites will send blood to this
central laboratory via the Department of Psychiatry and Psychotherapy at the
Klinikum der Universität München. If the reference ranges of
the trial drugs have not been reached and patients have not obtained remission
criteria (assessed at visits 5 and 7), the dose should be increased to reach
blood levels within the recommended range. To avoid unblinding, investigators,
patients and raters will only receive blinded qualitative information about the
drug levels. Thus, the rater will be informed whether the blood level of the
patient is within the therapeutic range or not and the actual blood level of the
investigational drug will be graded according to the following ranges:
below (<) the recommended therapeutic range
lower third within the recommended therapeutic range
middle third within the recommended therapeutic range
upper third within the recommended therapeutic range
above (>) the recommended therapeutic range, but below the
warning threshold
warning threshold
The quantitative information of the drug levels will be stored blinded at the
laboratory and included post hoc in the study database after the end of the
study for further statistical analyses. Investigators will only receive the
descriptive information detailed above.
Endpoints and safety measures
The primary endpoint is the relative frequency of patients in remission
after 8 weeks according to the RSWG (Remission in Schizophrenia Working Group)
criteria [33 ], without the time criterion
[34 ]
[35 ]. The RSWG criteria [33 ]
link DSM-IV symptoms of schizophrenia with items of the PANSS [24 ]. As detailed elsewhere [33 ]
[36 ]
[37 ], three symptom clusters
need to be considered: (1) psychoticism/reality distortion (PANSS items:
delusions, unusual thought content and hallucinatory behaviour), (2)
disorganisation (PANSS items: conceptual disorganisation and
mannerisms/posturing), and (3) negative symptoms (PANSS items: blunted
affect, social withdrawal, lack of spontaneity). According to the RSWG criteria,
the symptomatic criterion states that to achieve symptomatic remission, all
items must be rated as absent or present only to a mild degree (PANSS
value≤3).
Secondary endpoints are the relative frequency of patients in remission
after 4 weeks (early remission) according to the RSWG criteria [33 ], the change in PANSS total and in the
three PANSS subscales (positive, negative, general) from baseline to week 4 and
week 8, the frequency of patients in remission according to the RSWG criteria
without the negative symptom Items (N1, N4 and N6) after 4, 6 and 8 weeks, and
the frequency of patients with a clinical response according to PANSS
(≥20% reduction from baseline, corrected PANSS formula) [35 ] after 4 and 8 weeks. Safety
secondary endpoints include the change in white/complete blood
count (WBC/CBC), creatinine kinase (CK), blood levels from screening to
every visit during the study period (until visit 13), the relative change in
Troponin, frequency of 2-fold elevated Troponin, and absolute change in
C-reactive protein (CRP) values from screening to every visit during the first 4
weeks of the intervention period and at week 6 and week 8 as well as changes in
standard parameters of Electrocardiography (ECG) (QT interval, heart rate value
from screening/baseline to week 2, week 4, week 6 and week eight).
Several other assessments have been a priori-defined as detailed in [Table 3 ]. These other assessments include
among others the relationship between drug blood levels/concentration
(see below) and clinical and side-effect endpoints, changes in the Thought and
Language Disorder questionnaire (TALD) [38 ], the Clinical Global Impression scale (CGI) [39 ], the Personal and Social Performance
scale (PSP) [40 ], the Global Assessment of
Functioning scale (GAF) [41 ], the Calgary
Depression Scale for Schizophrenia (CDSS) [42 ], the Trail-Making Test [43 ], the InterSePT scale (ISST) [44 ], the abbreviated quality of life enjoyment and satisfaction
questionnaire (Q-LES-Q-18) [45 ], the Short
Form (12) Health Survey Change (SF-12) [46 ], the Drug Attitude Inventory (DAI10) [47 ], the Subjective Wellbeing under
Neuroleptics short form (SWN-K) [48 ] scale
and the attitude of patients towards participation in clinical research and
towards the treatment with olanzapine and clozapine at baseline (self-developed
questionnaire, in collaboration with a relatives-driven pan-organisation that
represents the interests of relatives and persons affected by mental
illness).
Other safety assessments include measurement of vital signs (change in heart
rate, blood pressure), change in metabolic parameters (weight and BMI), change
in extrapyramidal symptoms according to the Glasgow Antipsychotic Side-effect
Scale for Clozapine (GASS for Clozapine) [49 ] and the St. Hans Rating Scale (SHRS) [50 ], change in treatment-associated
constipation according to the Cleveland Clinic Constipation Score (CCCS) [51 ], change in akathisia according to the
Barnes Akathisia Rating Scale (BARS) [52 ],
changes in the study laboratory (fasting glucose, cholesterol, HDL), changes in
the number of cigarettes per day and changes in standard parameters of
Electroencephalography (EEG) (see [Table
3 ]).
Adverse events (AE), severe adverse events (SAE) and suspected unexpected serious
adverse reactions (SUSAR) are documented following established definitions and
legal requirements. The intensity of AEs is defined according to the common
terminology criteria for adverse events (CTCAE Version 4.0,
https://www.eortc.be/services/doc/ctc/CTCAE_4.03_2010–06–14_QuickReference_5×7.pdf).
In the event of a medical emergency, there are unblinding envelopes for each
patient at the respective study center.
During the safety follow-up period (starts after completed visit 9), only
laboratory-related AEs and SAEs will be documented.
A patient-relatives organisation (Bavarian Association of the Relatives of
Mentally-ill Patients (LApK)) is involved in the study. The organization
participates in the application process, assists in investigating the
patients’ view on treatment and serves as a consultant for study
patients. The aspect of the view of patients on treatment is especially
important because some studies have found that psychiatrists and patients have
different valuations of clozapine side effects [53 ], and one Cochrane review states that former trials have neglected
patients’ attitude towards clozapine [54 ].
Treatment after the end of the double-blind treatment phase
Unblinding patients at the end of the double-blind treatment phase (until visit
9) is sensitive, as it is not acceptable that the advantages of a double-blind
study are diminished by the possibility of changing previously collected data
after the treatment is unblinded. Therefore, a standard operating procedure
(SOP) was implemented on the means to continue treatment after the end of the
double-blind treatment phase. After visit 9 (day 56), the physicians being
responsible for the further treatment receive an envelope with the information
in which arm the patient has been allocated (unblinding according to protocol).
People involved in the trial will not receive this information. For patients who
drop out prior to visit 9, unblinding according to the protocol will also only
be possible following the aforementioned principles.
For patients who meet the remission criteria at week 8, it is recommended to stay
on the same drug. For patients who do not meet remission criteria at week 8 and
being treated with clozapine, it is recommended to carry on treatment and
perhaps increase the dose as clinical knowledge is available that the effect of
clozapine may occur with a delay. For patients being treated with olanzapine not
meeting remission criteria, it is recommended to increase the dose and, if not
successful, switch to clozapine.
However, the decision of any further treatment after visit 9 will be made between
the treating physicians and the patients following national guideline
recommendations. The study team will not be involved in this decision. Clozapine
should not be stopped abruptly but tapered down according to clinical standard
procedures. Therefore, our SOP shall assure that patients at the end of the
double-blind treatment phase will have enough study medication until the next
meeting with their treating physician while aiming at coordinating the meeting
with the treating physician on the same day of visit 9.
Sample size justification and planned data analysis strategy
The sample size calculation was performed on the basis of previous studies,
meta-analyses and systematic reviews [37 ]
[55 ]
[56 ]
[57 ], which indicate remission rates of approximately 30%
after 4 weeks [34 ] for antipsychotics in
acute schizophrenia and mean remission rates of ~ 50% after 8 to
12 weeks. Because olanzapine is one of the most effective antipsychotics [26 ]
[58 ]
[59 ], we proposed a
remission rate of 55% in the non-clozapine group after 8 weeks. Assuming
that 55% (=p2) of acute schizophrenia patients will remit after
treatment with olanzapine and assuming a superiority of clozapine in the range
of 10% to 30% in achieving remission [60 ] in severely affected patients, we
expect that clozapine will be associated with a≥20% (=p1
– p2) increased likelihood of achieving remission compared to
non-clozapine treatment. A two-group Chi² test (α=0.05,
two-sided, power [1-β]: 80%) will detect a difference between
groups after adjusting p1 (clozapine) to 0.74 and p2 (olanzapine) to 0.56 (odds
ratio of 2.236) when the sample size in each group is 110 (computed by nQuery
Advisor 7.0). Thus, the total required sample size is 220. The adjustment of
relative frequencies is due to the conservative assessment of the primary
outcome as outlined below, assuming less than 2% losses to follow-up
with no primary outcome data. As we aim to evaluate the primary outcome in every
participant, irrespective of whether he or she discontinues treatment, the
estimated rate of 2% losses to follow-up/drop-outs defines a
realistic scenario.
A statistical analysis plan will be developed before any analyses will be
performed. The statistical analyses will be performed after the database hard
lock (end of study), but before the end of the naturalistic follow-up. The
latter is a secondary analysis that will be performed at the end of the
naturalistic follow-up period. Descriptive statistics will be provided for all
data broken down by treatment group, visit and by country. Mean, median,
standard deviation, range, interquartile range and number of observations will
describe continuous variables. Absolute and relative frequencies will describe
categorical variables. All statistical tests will be carried out in two-tailed,
with the alpha (level of significance) being 5%. The randomization
procedure will be evaluated by a comparison of the two treatment groups for all
relevant variables recorded at baseline. The primary efficacy endpoint in the
intention-to-treat (ITT) population will be compared between groups by a
confirmatory Mantel-Haenszel test, with centres as strata on a two-sided
significance level of 5%. The analysis will be performed on the (ITT)
population, which will include all patients as randomized. To ensure a
conservative assessment of the primary outcome, losses to follow-up in the
experimental group will be rated as failures, while those in the control group
will be rated as successes. This approach is designed to avoid a beneficial bias
for the experimental group, which might be induced by non-compliance and losses
to follow-up. An additional subgroup analysis will be performed by logistic
regression using the primary efficacy endpoint as the outcome and the factor
variables group, centre and the categorized number of previous lifetime
treatments (cNLT categories: 0, 1,>1) as covariates. An interaction
effect between the group and the cNLT subgroups will be included in the model.
Parameter estimates of the model will be used to obtain point and interval
estimates (95% confidence intervals) of the treatment effect,
represented by Odds Ratios, within cNLT subgroups. The hypothesis test on the
model’s interaction effect equals the test for differences in the
treatment effects between subgroups. This subgroup analysis is exploratory and
none of the hypothesis tests of effects within and between subgroups is powered
by the study’s sample size calculation. The major objective is effect
estimation instead. Depending on the data distribution, group comparisons of
continuous and categorical secondary endpoints will be performed by linear or
binary logistic regression analysis that includes a factor variable for centres.
Baseline values will be included as another covariate if existing. In the case
of non-normally distributed residuals of the linear model, a van Elteren test
(=stratified Wilcoxon test) with centres as strata will be applied in an
additional analysis. Corresponding descriptive statistics and confidence
intervals will also be given. Absolute and relative frequencies of categorical
safety outcomes will be assessed along the course of the trial (safety
population). Differences between groups will also be tested for statistical
significance with Fisher’s exact test. Likewise, continuously
distributed safety outcomes (safety population) will be presented by descriptive
statistics and compared with t-tests or Mann-Whitney U tests, as appropriate.
All tests of secondary endpoints will be computed in the ITT and per protocol
(PP) populations in an explorative manner on two-sided significance levels of
5%. Exploratory analyses will be performed as appropriate for other
assessments.
The ITT population includes all patients as randomized. The PP population will
include all participants without major protocol violations. The safety
population will include all participants who received at least one study drug
and will be analysed as treated. The allocation of patients to the respective
study populations will finally be determined in a blinded data review meeting
(BDRM). All analyses will be performed in accordance with the ICH Guidelines
E9.
Discussion
An early clozapine application, as proposed in the EARLY trial in disease stages
after the first episode and before treatment-resistance (second-line treatment), has
the potential to increase remission rates and to prevent poor disease outcomes and
chronicity in schizophrenia. Poor remission rates and frequent relapses in
schizophrenia contribute to the highest direct costs of all brain disorders in
Europe, totaling 29 billion Euros per year [62 ]. Furthermore, early disease chronicity entails physical and mental
comorbidity, enduring social and vocational exclusion and excess mortality, which
alone in Europe amount to annual indirect costs of 65 billion Euros [62 ].
Multiple sources of evidence show that schizophrenia patients have a high likelihood
of response during their first episode [1 ] but
that response rates drop after the first relapse. In a small naturalistic study,
38% of first-episode patients were non-responsive to first-line treatment
with risperidone and among those non-responders, only 44% achieved full
remission of positive symptoms after treatment with olanzapine (second-line) [63 ]. As many patients do not remit after the
first relapse and as remission is related to recovery [64 ], good functional outcomes and improved
quality of life [56 ], new evidence-based
strategies are urgently needed to increase remission rates. An increase in remission
will not only improve the aforementioned outcome domains but is likely to reduce the
risks of hospitalization and treatment discontinuation. In this context, the early
application of clozapine, the most effective antipsychotic, has been discussed now
for years as one possibility to improve outcomes in relapsing schizophrenia [12 ]. This issue of a potential indication
extension of clozapine from being the ‘last-resort antipsychotic’ to
being applied at an earlier stage is supported by several lines of evidence [12 ], but respective clinical trials to test
these hypotheses are lacking. To date, sufficiently powered clinical trials
investigating the proposed superior efficacy of an early clozapine application
compared to standard treatment are lacking. A retrospective cohort study confirmed
the hypothesized superior effectiveness of clozapine compared to standard
antipsychotics with regard to time to hospitalization and risk for treatment
discontinuation [65 ], and clozapine is still
the gold standard in treatment-refractory cases [16 ]
[17 ]
[18 ]. Moreover, one nationwide cohort study
investigated the real-world effectiveness of 29823 patients with schizophrenia and
showed the association between clozapine treatment and the lowest rates of treatment
failure compared to all other oral antipsychotics [66 ]. In contrast to guideline recommendations [16 ]
[17 ]
[18 ], in clinical practice,
clozapine use is delayed by many years, making clear that more convincing evidence
is needed to overcome this discrepancy [20 ].
We hypothesize that the early introduction of clozapine in patients with acute
schizophrenia will result in higher remission frequencies and thus reduce the risk
of developing unfavorable disease outcomes. In summary, there is broad evidence that
clozapine is an effective antipsychotic that can be safely used in
non-treatment-refractory patients and that has superior effectiveness in real-world
settings. Clinical trials investigating the role of clozapine in acute relapsing
schizophrenia are lacking (second-line treatment). Such an approach also must
specifically address the potentially high levels of side effects and AEs to allow
for a comprehensive risk-benefit evaluation. An increase in the likelihood of
remission has the potential to prevent treatment-resistance and long-standing
disability in schizophrenia patients with a relapsing disease course and thus reduce
the overall socioeconomic burden. Because neither study drug is under patent
protection, no commercial interest for any pharmaceutical company exists. Therefore,
this trial has the potential to change guideline recommendations in an area where
no
new development from the perspective of the pharmaceutical industry can be
expected.
Moreover, this trial will also allow investigating whether low dosages of those
highly effective antipsychotics are sufficient to reach symptomatic remission
following the discussions of minimal effective dose treatment for an optimal
risk-benefit evaluation. The latter will be possible because the EARLY trial
includes the so far most extensive side-effect evaluation of clozapine and
olanzapine in a controlled design going far beyond the safety assessment in clinical
practice.
Impact of SarsCov2-Pandemic (Covid-19)
Specific measures were implemented to increase patient and data safety while
conducting the trial during the pandemic. A specific risk-based approach dealing
with the challenges of the pandemic based on the recommendations laid down in
the EMA guidance on the management of clinical trials during the covid-19
(coronavirus) pandemic [67 ] the
recommendations of the Federal Institute for Drugs and Medical Devices (BfArM)
[68 ] and the recommendations of the
Work Group of Medical Ethics Commissions in Germany for clinical research during
the pandemic [69 ]. Furthermore, the
planned period for patient recruitment had to be substantially prolonged due to
restrictions and the impact of the SarsCov2-Pandemic.
