id
stringlengths 36
36
| text
stringlengths 95
399
| source
stringclasses 1
value |
|---|---|---|
a3d00891-02ae-42ce-b86b-32a85db75811
|
with acute traumatic brain injury,34 and adults with COVID-19.11, 12, 17, 26, 30, 44, 45, 53, 66
Utilizing a sparse-sampling approach, we will measure anakinra concentrations by enzyme-
linked immunosorbent assay in blood samples of subjects who are randomized to study drug.
During each immune function testing, one blood sample per subject will be obtained. We
|
protocol.txt
|
e8a548df-d119-4f58-b087-9ff187dbbc97
|
anticipate that this sampling will occur in subjects at various times during the dosing interval. As
such, this sparse sampling strategy will provide adequate data for a population pharmacokinetic
modeling approach that will describe anakinra pharmacokinetic parameters in critically ill
children and evaluate the effects of organ dysfunction and systemic inflammation on drug
disposition.
|
protocol.txt
|
115fc665-8c48-44ac-9223-305687f6dc4d
|
4.4.3 Justification for Anakinra Dosing Strategy
The dose of anakinra that might be effective in improving outcomes in hyper-inflamed children
with sepsis-induced MODS is unknown. Doses at or above the highest FDA-approved dose
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
|
protocol.txt
|
e60d85b0-3e36-4dce-a4d2-74399d70ae15
|
The Collaborative Pediatric Critical Care Research NetworkPage 32 of 76 Protocol 90 (Hall, Zuppa and Mourani)
(8mg/kg/day) have been frequently reported to be required for symptom control in children with
inflammatory syndromes.14, 41, 44, 45, 65, 67 In the 1994 and 1997 adult sepsis trials, a dose of 48
|
protocol.txt
|
aef2e0a9-3d5b-4319-84da-3f314cb13e69
|
mg/kg/day given for three days by continuous infusion was used.23, 63 Despite its apparent safety
in adults, this dose is substantially higher than what has been used in children with HLH/MAS
or adults with COVID-19.
In the largest (to date) published study of anakinra in hospitalized, hyper-inflamed adults
with COVID-19 (N=392), a dose of 10 mg/kg/day IV divided every 12 hours (infused over 1
|
protocol.txt
|
dddd5f5d-d037-4159-ab82-765e8f50e175
|
hour) was associated with lower mortality compared to contemporaneous controls (hazard ratio
0.45, 95% CI 0.20 – 0.99, p=0.047).12 This was in agreement with a prior pilot study (N=45)
that showed lower mortality in the same population with this dosing strategy (10% vs 40%,
p=0.009).11 No drug-attributable adverse events (including no increase in risk of nosocomial
|
protocol.txt
|
6c6aff87-21b5-45c3-8b0d-52d15f1eb395
|
infection) were reported in these studies. The Cavalli studies11, 12 (10 mg/kg/day) suggested
beneficial effect in adults with COVID-19 who had elevations in serum ferritin or CRP levels
similar to those targeted in the TRIPS trial. The adult sepsis trials showed similar benefit of
anakinra in subjects with severe inflammation, but it is not clear if doses of 48 mg/kg/day will
|
protocol.txt
|
5a8d6437-5dca-49dd-ba7e-3781e0cf41e1
|
be safe in the pediatric population, nor is it clear that doses that high are necessary to improve
outcomes from pediatric sepsis-induced MODS.
The optimal duration of anakinra therapy for hyper-inflamed children with sepsis-induced
MODS is similarly unknown. The adult sepsis studies utilized a continuous infusion over 72
|
protocol.txt
|
426b1ec3-b9da-4d67-be75-c463cc135799
|
hours (for a total of 144 mg/kg). Those studies targeted the acute phase of sepsis (within 24
hours of sepsis onset) whereas the TRIPS trial targets the subacute phase of sepsis-induced
MODS. Our prior work with immunomodulation in pediatric MODS suggests that 7 days of
treatment is effective in normalizing immune function,32 and the duration of anakinra treatment
|
protocol.txt
|
458f385c-d01c-4203-9086-689098fff610
|
reported in adult COVID-19 studies has ranged from 5 days to 14 days (where treatment duration
is specified).
The TRIPS trial will therefore adopt an adaptive dosing strategy that includes the FDA-
approved dosing range and extends it to include doses that have been associated with benefit in
HLH/MAS and COVID-19. Subjects who meet criteria for receiving study drug in the TRIPS
|
protocol.txt
|
68b6e740-cb4c-4d75-baef-f4f2678f0119
|
trial will be adaptively randomized to receive placebo or anakinra at a dose of 4, 8, 12, or 16
mg/kg/day IV divided every 12 hours (+/-2 hours from the previous dose), each dose infused
over 1 hour, for 7 days. Initial randomization allocation will be skewed toward lower doses,
with increasing proportions of enrollments being allocated to higher doses over time if there is a
|
protocol.txt
|
732165a5-1d76-415b-84ad-3a931005f842
|
favorable safety profile with lower doses and a suggestion of benefit with higher doses.
The q12 hour dosing approach was chosen for the following reasons:
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkProtocol 90 (Hall, Zuppa and Mourani) Page 33 of 76
|
protocol.txt
|
a7a62ca8-e241-42eb-89d3-54b719329d58
|
• Due to a lower number of available IV lumens in critically ill children (compared to
adults), and with a relative lack of drug compatibility data, it would be unfeasible to
occupy a dedicated IV line for 7 days. Intermittent dosing removes these barriers.
• We conducted pharmacokinetic simulations on bolus vs infusion dosing regimens, using
|
protocol.txt
|
14539d35-6c95-402b-83e3-f1abadb9c4dc
|
published pharmacokinetic data (allometrically scaled for pediatrics), and found that
IV administration over 1 hour is predicted to result in a Cmax across our dosing range
that is similar to those seen in the adult sepsis trials, but with a similar area under the
concentration curve (AUC) to IV bolus dosing (in which Cmax would be substantially
|
protocol.txt
|
1918daa8-760c-42f5-9e4a-52cdb42438f6
|
higher). The predicted Cmax with a 1-hour infusion are similar to those seen in adult
studies in which there were no drug-attributable adverse events.
4.4.4 Dose Adjustment for Severe Renal Dysfunction
Anakinra is eliminated through both receptor-based clearance and renal clearance.84 The
package insert recommends a 50% dose reduction (switching from daily to every-other-day
|
protocol.txt
|
5f15895e-f618-4c02-9ea8-e97651c74b55
|
dosing) in patients with a creatinine clearance (CrCL)<30 ml/min. We will therefore plan a
50% dose reduction in subjects with CrCL <30 ml/min. Our simulations predict that, in our
patient population with severe renal dysfunction, adopting a once-daily dosing approach using
50% of the assigned dose, will result in a Cmax and AUC that are similar to subjects with intact
|
protocol.txt
|
7bcca81a-3bba-4843-835f-f4c39161a071
|
renal function who are receiving the full assigned dose.
4.4.5 Pharmacy Function and Monitoring
The investigational pharmacy at Nationwide Children’s Hospital will serve as the central
pharmacy for the TRIPS study. They will intake anakinra from the manufacturer (Sobi),
document lot number and expiration information, facilitate blinding, and ship study drug
|
protocol.txt
|
8518f532-8e03-4d6f-96c9-a1fec47696ff
|
to clinical sites’ investigational pharmacies. They will also develop source documentation,
pharmacy training materials, operating manuals, and study-specific pharmacy procedures as
needed. Clinical sites’ investigational pharmacies will be responsible for preparing placebo
doses, all of which will consist of normal saline. The clinical site pharmacy must maintain
|
protocol.txt
|
50377f57-2054-47ab-87bc-bd3443a394a9
|
adequate records of all dispensed study drug. Each pharmacy will be monitored and may
be requested to send copies of these documents to the Data Coordinating Center and/or the
investigational pharmacy at Nationwide Children’s Hospital which will conduct remote audits
at regular intervals.
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
|
protocol.txt
|
b1a33ab7-2206-45b3-8103-f0a007333de7
|
The Collaborative Pediatric Critical Care Research NetworkPage 34 of 76 Protocol 90 (Hall, Zuppa and Mourani)
4.5 Randomization
Upon determination of a subject’s immunophenotype, Dr. Hall or his designee will notify the
Data Coordinating Center and the clinical site investigator of the laboratory results. Subjects
|
protocol.txt
|
6125836a-a53d-4520-a2c7-90b41218ea23
|
who are eligible to undergo randomization in the study will be randomized by the DCC to active
or placebo arms. The randomization system or DCC personnel will notify the investigational
pharmacy at the clinical site and the site’s investigational pharmacy will provide the proper
blinded study drug to the subject’s bedside for infusion. The DCC will also notify the study’s
|
protocol.txt
|
d2627a18-1f59-47d2-94f5-b87f4f07ffb1
|
central pharmacy at Nationwide Children’s Hospital of the randomization results so that the
central pharmacy can track and ensure drug supply.
4.6 Blinding Study Arms
The GRACE-2 trial is a double-blind trial. Subjects will receive GM-CSF or saline placebo
intravenously over at least six hours daily for seven days. The TRIPS trial is a double-blind
|
protocol.txt
|
9181b098-7ace-4b1f-8f2c-e7d8b22a7c4d
|
trial. Subjects will receive anakinra or normal saline placebo every 12 hours (+/- 2 hours from
the previous dose) for seven days. All active drug and placebo doses are intravenous. Because
varying doses of anakinra are being evaluated, the drug will be diluted to the equivalent volume
of the highest possible dose (16 mg/kg/day) by the investigational pharmacies prior to dispensing
|
protocol.txt
|
74763a2f-6118-4356-b6bb-cc5552611883
|
in order to maintain blinding. The saline placebo will be similarly dispensed in a volume to
match what would be the volume of the 16 mg/kg/day dose of anakinra.
4.7 Biorepository Sample Collection and Processing
The PRECISE study will include the serial collection of biological samples including serum
and plasma, RNA tubes, cell pellets for DNA, and endotracheal aspirates in order to answer
|
protocol.txt
|
edeea600-b3e8-42de-92a4-67d9875bb03b
|
mechanistic questions around the biology of immune function and inflammation in children with
sepsis-induced MODS. For the purposes of this submission, these samples are being referred to
as a “biorepository”, though all samples and analyses will be directly related to the PRECISE
study. They will permit better understand the pathophysiology of sepsis-induced MODS and
|
protocol.txt
|
f911f760-017e-49e6-a27e-1d6a003ea6bd
|
will help explain the results of the trial. These analyses may include quantitation of biomarkers
in serum or plasma, cellular phenotyping, microbiome analyses, quantitation of RNA expression,
and DNA analyses including sequencing.
A modest amount of blood (approximately 2 – 4ml) will be collected at each immune
|
protocol.txt
|
49e5ced3-7845-45de-a7f8-5771c475e725
|
phenotype sampling time, in addition to the volume required for immune phenotyping, along
with an endotracheal aspirate (if an endotracheal tube is present). The total volume of blood
taken will be adjusted based on the subject’s weight such that the total study-related blood
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
|
protocol.txt
|
09ab9eb4-deca-4cf0-ba95-28e4460de75f
|
The Collaborative Pediatric Critical Care Research NetworkProtocol 90 (Hall, Zuppa and Mourani) Page 35 of 76
volume is <5ml/kg on MODS Day 2 and <9.5ml/kg over the entire study.
Parental Permission/Consent/Assent for the collection of these samples is included in the
procedures for the PRECISE study. Parental Permission/Consent/Assent allowing for additional
|
protocol.txt
|
86524d99-c6e4-4751-a0ac-5e7f3fe59960
|
uses of these samples, unrelated to the PRECISE study, will be incorporated into the PRECISE
parental permission/consent/assent form with opt-in/opt-out options. These will include an
option to only use residual samples other than DNA sequence data.
Biorepository samples will be shipped from sites quarterly to Dr. Hall at Nationwide Chil-
|
protocol.txt
|
3e2c59cd-446b-4513-ac25-4607e1f378c8
|
dren’s Hospital where they will be accessioned and stored. All samples will be deidentified prior
to shipment to Nationwide and will be linked to a Master List using a unique code which will
be managed by the DCC PI. Access to the specimens and linked clinical data will be governed
by policies set forth by the CPCCRN Steering Committee, including the DCC PI, who will
|
protocol.txt
|
52ab291f-318f-4e66-991d-4a00a614f98c
|
constitute the CPCCRN Biorepository Governance Committee (BGC) and will be responsible
for reviewing applications requesting biorepository samples.
Proposal critiques and/or suggestions for potential collaborative efforts (for overlapping
projects) will be provided. The policy for access to specimens will include the following
principles:
|
protocol.txt
|
a8f22cc2-f6e4-4376-8a73-cf4acd19d606
|
• highest priority will be given to externally-funded studies of high scientific merit from an
investigator who participated in the primary study;
• overlapping proposals that duplicate specimen testing will not be approved;
• investigators will demonstrate that they can perform the stated experiments and have
obtained local IRB approval to use biorepository specimens;
|
protocol.txt
|
b977efd6-eeb0-4707-a565-1e35bbd01a9d
|
• for rejected requests, the BGC will provide the investigator with a detailed reasoning for
rejection.
4.8 Discontinuation of Study Drug
Study drug may be discontinued if an adverse event occurs that requires discontinuation in the
opinion of the clinical or investigative teams, if a contraindication to active drug develops, or
|
protocol.txt
|
23969d27-0a4c-4cc0-8831-63123609d466
|
if parents request discontinuation of study drug. This does not constitute withdrawal from the
study, and all data and sample collections should continue.
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkPage 36 of 76 Protocol 90 (Hall, Zuppa and Mourani)
4.9 Withdrawal from Study
|
protocol.txt
|
22b9b1a3-5deb-46a4-875b-3e05cc66b659
|
Parents may withdraw consent for their child to be in the study at any time, but such an
occurrence should be very rare. In most instances, if a family or subject indicates that they
want to stop being in a study, they usually mean they wish to stop the intervention and/or blood
sampling. Efforts should be made to be able to continue all data collection and sampling. In the
|
protocol.txt
|
2d002b9a-62bf-446b-800f-a81af8883e60
|
event that the parents insist on cessation of data collection and sampling, all study procedures
and data collection will be discontinued, though we will continue to review study-related data
collected prior to the subject’s withdrawal from the study.
4.10 PK Sample Collection
The goals of PK sampling are to understand how the host immunophenotype and organ dysfunc-
|
protocol.txt
|
2800ffc8-61f9-48d3-9b5b-3081c4bab64f
|
tion in sepsis influence antibiotic concentrations over time, and how antibiotic concentrations
and the host immune response impact clinical and microbiological outcomes. A subset of
participating sites, chosen for their infrastructure, will collect up to 15 extra blood samples
for the purpose of measuring concentrations of the following antibiotics: vancomycin, cef-
|
protocol.txt
|
55cd4435-a9bf-4ef4-9105-b4be22938df1
|
triaxone, cefepime, meropenem, piperacillin/tazobactam, ampicillin, or ampicillin/sulbactam.
Only subjects who are receiving one or more of these antibiotics will have the extra blood
samples collected. In conjunction with each immunophenotype sampling, up to 3 samples (20
µL/sample in duplicate) will be collected for quantification of antibiotic concentrations in whole
|
protocol.txt
|
7eaeb6cf-839a-406b-95cf-6f21d61d3edc
|
blood (maximum 600 µL for the entire study). Sampling will take place as close as possible to
immunophenotype sample collection and may be performed using an existing arterial or venous
catheter, venipuncture, or capillary stick. After collection, samples will be dried and then frozen
locally at -80◦ C within 72 hours of collection. They will be analyzed at a central laboratory.
5 Data Collection
|
protocol.txt
|
69e3bd99-4278-4141-8025-7a012891cc84
|
The schedule of activities and data collection is shown in Table 2 on page 39. Data elements to
be collected are briefly described below. The described data are not intended to be inclusive,
and additional data that are similar in nature may be added to the final study database without
revision of this protocol. If there are entirely new areas of data collection added during the
|
protocol.txt
|
778347eb-9f0b-4ba1-98a0-ade810aa938e
|
study, a protocol amendment will be filed with the sIRB.
5.1 Demographic Data
Demographic data may include birthdate, gender, race, ethnicity, and socioeconomic status.
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkProtocol 90 (Hall, Zuppa and Mourani) Page 37 of 76
5.2 Eligibility Data
|
protocol.txt
|
a668a889-d483-4f98-ab14-204d8b34901e
|
For all screened patients who meet inclusion criteria, the inclusion and exclusion criteria will
be entered, as well as the final eligibility status. Exclusion criteria are recorded to create the
eventual CONSORT diagram as well as to inform decisions about potential protocol revisions
if the exclusion criteria are determined to be too restrictive. Changes in exclusion criteria will
|
protocol.txt
|
761625a6-55e7-41db-8c7f-f19537512aab
|
require sIRB, DSMB and FDA approval before implementation.
5.3 Consent Procedure Data
For all eligible patients, we will record whether the parents were approached for study consent,
whether they consented, and date of consent, if applicable. In addition, reasons for not approach-
ing parents of eligible subjects will be recorded, and if parents are willing to express a reason
|
protocol.txt
|
d8d17eb0-5394-41d9-9330-8cbc13c2a0ce
|
for not consenting, this will also be recorded.
5.4 Baseline Review of Systems
Medical historical data and review of systems will be recorded to provide a baseline against
which to evaluate adverse events.
5.5 Baseline Admission Data
Dates of admission to hospital and PICU, reasons for hospital admission, co-morbidities,
allergies, height, weight, and physical findings will be recorded.
|
protocol.txt
|
24bf0cd1-b8ba-46ec-a57b-19f3da48b976
|
5.6 PICU/WARD Daily Data
Data concerning the physiologic status, therapies, and significant events will be recorded for
each day until 28 days after randomization, hospital discharge, or death, whichever occurs first.
5.7 PELOD-2 Data
Data required for calculation of the daily PELOD-2 score will be collected.
5.8 Immunophenotyping
|
protocol.txt
|
11d8ddf7-3d21-43f1-8b62-65374b1a6233
|
Data collection will include the time of sampling, relevant times involved in sample processing,
temperature maintenance, and shipping details.
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkPage 38 of 76 Protocol 90 (Hall, Zuppa and Mourani)
5.9 Randomization Procedures
|
protocol.txt
|
c722e238-dc0e-4a47-aac5-83eb58a96041
|
Randomization data will include information provided to the randomization system, confirmation
of the phenotype and eligibility for interventional study, information required for stratification
of randomization, and the actual randomization number received from the system and assigned
to the subject.
5.10 Drug Administration Data
|
protocol.txt
|
b4581437-ecf0-4f04-867c-40d29d56c01d
|
Drug administration data include time of administration, dose of study drug, and whether there
were any adverse reactions noted during the administration.
5.11 Biorepository Sampling
Data will include the date and time of collection, which biospecimens were obtained, cold
storage and processing information, and shipping information. Samples should be obtained at
|
protocol.txt
|
87d6219a-9a57-4a32-a2e7-0e2a44901c79
|
baseline, and on the days of immunophenotyping sampling. If blood volume becomes limiting,
biorepository sampling may be reduced.
5.12 Quality of Life Instruments
Instruments to assess aspects of quality of life include the PedsQL generic scales or infant
scales, the Functional Status Scale, the PedsQL Family Impact Module 2.0 (FIM) and the
|
protocol.txt
|
b2c2f4e1-430f-4a9a-8221-bfd241a773fc
|
Pediatric Evaluation of Disability Inventory - Computer Adaptive Test (PEDI-CAT). These will
be collected as early as possible for baseline assessment, and at 3 (allowable interval 3 to 5
months) and 12 months (allowable interval 11 to 13 months) after MODS Day 2.
5.13 Data Logs
Data forms will be created to collect concomitant medications, laboratory data that are not
|
protocol.txt
|
30f1695d-2663-429f-8f55-12eba485c654
|
collected on daily forms, and adverse events.
5.14 Discontinuation or Withdrawal from Study
If a subject is discontinued from study drug or withdrawn from study, data concerning the event
will be recorded.
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkThe Collaborative Pediatric Critical Care Research Network
|
protocol.txt
|
f7b363d7-1414-4e64-83f7-dd81529c1472
|
Protocol Version Date: June 16, 2023
PRECISE Protocol Version 1.07
Definitions of Study Days
Study Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 to 28 3 Months 12 Months
MODS Day 0 1 2
Not defined - MODS may resolve by day 3.
Enrollment window
Study Drug Day* Not applicable
1 2 3 4 5 6 7
Not applicable
Demographic Data X
Eligibility Data X
Consent Procedures X
Baseline Review of Systems X
|
protocol.txt
|
fa81917c-31aa-4586-8d9a-09724540e179
|
Baseline Admission Data X
PICU/WARD Daily Data X X X X X X X X X X X X X X X X X
PELOD-2 Data X X X X X X X X X X X X X X X X X
12 +/- 1d
Immunophenotyping +/- PK X X X 17 +/- 1d
Randomization Procedures* X
Drug Administration Data* X X X X X X X
12 +/- 1d
Biorepository Sampling X X X 17 +/- 1d
PedsQL X X
X (Baseline Status)
FSS X X
X (Baseline Status)
PedsQL FIM 2.0 X X
X (Baseline Status)
|
protocol.txt
|
ee3c358d-615f-4f85-93dd-a28fe785bc0c
|
PEDI-CAT X X
X (Baseline Status)
Readmissions X X
Concommitant Medications Log X X X X X X X X X X X X X X X X X
Laboratory Data Log X X X X X X X X X X X X X X X X X
Adverse Events Log* X X X X X X X X X X X X X X X X
Discontinuation of Study Drug*
Withdrawal from Study*
Table 2: Schedule of study events and data collection. Asterisks indicate When applicable.
|
protocol.txt
|
b92299b1-daaa-4b1b-9d1a-5a2250135835
|
Protocol 90 (Hall, Zuppa and Mourani) Page 39 of 76Page 40 of 76 Protocol 90 (Hall, Zuppa and Mourani)
6 Statistical Summary
6.1 GRACE-2 Randomization and Power Estimation
GRACE-2 trial’s “Goldilocks” framework85 uses the current observed distribution of the primary
outcome (sum of PELOD-2 over 28 days, with deaths assigned the worst possible score) for in-
|
protocol.txt
|
1e98c4d5-1c70-49db-b368-6aba1fa352a5
|
terim decision-making. Prespecified, trial-customized stopping cutpoints are used for declaring
significant treatment efficacy, or declaring futility of trial continuation, if compelling evidence
from the trial data to date indicate either action. Frequent interim efficacy data review yields
potentially earlier trial stopping (for efficacy or futility) than traditional frequentist monitoring
|
protocol.txt
|
88519acd-588d-43f8-add6-a7c4e8fcaae5
|
schemes.
The assumed placebo distribution for the primary outcome is based on the historic distri-
bution of PELOD-2 over 28 days from our published LAPSE study.86 GM-CSF is assumed to
improve mean sum of PELOD-2 over 28 days among survivors by approximately 20 points.
Figure 5 on the next page shows arm-specific histograms and quantiles. Mortality is assumed to
|
protocol.txt
|
0e32efe7-cd9e-4fea-b06b-e8d19c2d3027
|
decrease from 12% placebo to 9% in the active arm. Under this scenario, simulation-estimated
power with 400 randomized patients is 91% to detect a difference between GM-CSF and placebo,
using a one-sided Wilcoxon Rank Sum test to compare arms and controlling Type I error at
2.5%. The Statistical Design and Power document has more details on the adaptations and
|
protocol.txt
|
573aded1-754a-4556-b0c3-9a9c804a03eb
|
assumptions, as well as our design’s complete operating characteristics under this scenario and
others considered.
In the GRACE-2 trial, if we do not stop the trial early for efficacy or futility, the final
one-sided p-value threshold for a conclusion of success is 0.020.
6.2 TRIPS Randomization and Power Estimation
|
protocol.txt
|
8517f2f2-4a1a-4f9a-a676-71dbadefbd34
|
This study will utilize response adaptive randomization which updates allocation proportions
throughout the trial to preferentially allocate patients to arms appearing superior based on accu-
mulating data.80 We will utilize the “Goldilocks” principle of frequent looks at the accumulating
data to frequently alter the randomization proportions often and stop when the sample size is
“just right”.
|
protocol.txt
|
35296437-f07a-4cc8-b890-b8e6cd961bc8
|
Response adaptive randomization is used to allocate more patients to an active arm that
appears to be superior based on the accumulating data. The proportion of patients allocated to
placebo will be fixed throughout enrollment (described in more detail below). Once response
adaptive randomization begins, patients will be randomized in block sizes of 30 among the four
PRECISE Protocol Version 1.07
|
protocol.txt
|
686a39e8-c52d-4f55-ba6e-4033814f7d9a
|
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkProtocol 90 (Hall, Zuppa and Mourani) Page 41 of 76
Figure 5: Assumed distributions of GM-CSF and placebo for mortality and sum of PELOD-2
among survivors.
doses of anakinra (4 mg/kg/day, 8 mg/kg/day, 12 mg/kg/day, 16 mg/kg/day). The blocks will be
|
protocol.txt
|
0868a50a-fea4-4e13-8713-91d0776a6b2d
|
re-defined after each 30-patient interval starting at 180, based on the current estimates of the
Emax dose response curve. All slots in the block will be allocated adaptively. The proportion
will be determined based on simulation-based operating characteristics. At each interim, we will
calculate pr(ED85) for each arm using the currently available outcome data. To avoid assigning
|
protocol.txt
|
cd83cd53-657b-4247-85a1-40282d60323f
|
patients to an arm with a minimal chance of being the best dose, any allocation probability
less than 0.05 is set to zero at that interim and the resulting probability is reallocated among
the remaining arms. In this manner, an anakinra dose may be temporarily dropped but may
be re-introduced if the response adaptive randomization probability increases at subsequent
|
protocol.txt
|
5b441f33-3c70-43a8-816c-42387a9c93e6
|
interims. The placebo dose will never be dropped.
Throughout the study, 50% of patients will be allocated to placebo regardless of what
happens in the response adaptive randomization. This proportion will be chosen based on the
optimization of the operating characteristics in simulations. In order to establish an initial
|
protocol.txt
|
cf06faff-eaae-4fca-9124-62e46dc386c2
|
safety profile, a larger proportion of patients will be initially allocated to lower anakinra doses
(i.e., 4 mg/kg/day and 8 mg/kg/day) compared to the higher doses (i.e., 12 mg/kg/day and 16
mg/kg/day). After the first 180 patients are enrolled and a full safety review by the DSMB of
each dose, the randomization probabilities will be statistically updated based on the response
|
protocol.txt
|
e021fc56-8452-4e73-ad92-9adbae4c1a22
|
adaptive randomization described above. The goal is to allocate more subjects to the efficacious
arms while minimizing the allocations to higher doses when lower doses are almost equivalent
in efficacy. The allocations will be updated based on the collective study data. Allocation
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
|
protocol.txt
|
bfb48b1d-1b10-43a1-a530-0568195dbf00
|
The Collaborative Pediatric Critical Care Research NetworkPage 42 of 76 Protocol 90 (Hall, Zuppa and Mourani)
proportions will not be stratified by site nor any other clinical variable. Additional details can
be found in the Statistical Analysis Plan.
We have studied the immunology of pediatric sepsis-induced MODS in single- and multi-
|
protocol.txt
|
c57e66ba-9086-467b-826d-e7bc792dfc3c
|
center populations and have found the incidence of immunoparalysis to be 48-53% in children
who have sepsis-induced MODS using the Proulx criteria. This includes the CPCCRN network’s
GRACE-1 study (48%), Drs. Hall and Zuppa’s multi-center ongoing PARADIGM study (50% to
date), and Dr. Hall’s recent 102-patient single-center cohort of children with severe sepsis (53%).
|
protocol.txt
|
cfce6c15-e1a9-4f2f-9b20-ee41018f62ba
|
Among the 401 children with severe sepsis/septic shock in CPCCRN’s recently-completed multi-
center PHENOMS study (289 of whom had sepsis-induced MODS), 96% of the cohort had
a CRP >4 mg/dl and/or a ferritin level >500 ng/ml. Of those, 5% would not qualify for the
TRIPS trial by virtue of very severe hyperferritinemia. We anticipate that approximately 50% of
|
protocol.txt
|
13814381-0f67-421e-8d2b-0fe883a121b3
|
immunophenotyped patients will qualify for entry into the TRIPS trial. We plan to enroll up to
1095 patients into the overall study in order to accrue 500 patients into the TRIPS trial.
6.3 Data Analyses
6.3.1 Primary Outcome
The primary outcome is duration and severity of organ dysfunction as quantified by the cumula-
|
protocol.txt
|
dea516b7-ca15-4cf8-9805-e6aafde065b2
|
tive Pediatric Logistic Organ Dysfunction-2 (PELOD-2) score (the sum of each day’s score) at
28 days from time of randomization. On any given day, the minimum score is 0 and maximum
score is 33, with lower values indicating less organ dysfunction. This corresponds to a maximum
possible 28-day sum of 924. Mortality is incorporated into this analysis, with children who
|
protocol.txt
|
9d28ca03-525b-4e66-9e5c-b1c24a8d0500
|
die within 28 days of randomization assigned the worst possible score in a rank-based analysis.
Rank based analyses will be performed to account for mortality as well as the fact that for
surviving children, a difference in the cumulative PELOD-2 of a specified magnitude may reflect
different clinical consequences.
6.3.2 Secondary Outcomes
|
protocol.txt
|
50a7842c-850e-4d18-83c0-eeeb498518f7
|
The analytical plan is designed under the approach that 2.5% Type I Error (one-sided) is to be
allocated for the primary efficacy outcome comparison, while an additional (separate) 2.5% Type
I Error (one-sided) is allocated for efficacy assessment for a small number of fully prespecified
secondary efficacy outcomes.
|
protocol.txt
|
64cf12c4-5881-4349-8286-20282875e299
|
The two formally designated secondary efficacy outcomes, change in PedsQL from baseline
to 3 months and change in Functional Status Score (FSS) from baseline to 3 months, will be
evaluated for efficacy using all available data in the enrolled population of children. Significance
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
|
protocol.txt
|
f0c16025-38bc-439d-b9d3-475bdeca0821
|
The Collaborative Pediatric Critical Care Research NetworkProtocol 90 (Hall, Zuppa and Mourani) Page 43 of 76
of each these two outcomes will be evaluated using the same approach as the primary outcome
and the ranks will be modeled in order to account for deaths. Different non-informative priors
for the TRIPS analysis may be used to reflect the different scales. For the GRACE-2 analyses,
|
protocol.txt
|
71a3e464-5d8f-446c-8f11-0093a7eb3c30
|
we will implement a Bonferroni-Holm stepdown procedure to maximize power to detect a
significant effect while limiting the overall Type I error rate to 2.5% for these two comparisons.
6.3.3 Exploratory Outcomes
For the GRACE-2 trial, mortality rates and other binary outcomes will be compared between
treatment arms using chi-squared tests, stratified by site. MODS-free and organ-dysfunction-free
|
protocol.txt
|
e1dc329e-ebdb-4c33-bed4-4cb797bb3613
|
days will be compared between treatment arms in a rank-based fashion, accounting for site.
For the TRIPS trial, mortality rates, and other binary outcomes, will be compared between
the placebo and active arms using an Emax dose-response model with a logit link function. In
particular, the model for the primary outcome will be transformed such that the probability of
|
protocol.txt
|
c217f5fe-e176-4079-820d-f97bf17dc750
|
the event occurring is defined:
eθd
Pd =
1 + eθd
The remaining parameters and priors are the same except the variance parameter is excluded due
to the natural relationship between the mean and variance with the logit link. MODS-free and
organ-dysfunction-free days will be compared in a rank-based fashion similar to the primary
outcome.
|
protocol.txt
|
0d392aec-ec66-4113-bf40-4d64178fb7eb
|
For outcomes evaluated at both 3 and 12 months, we have elected to perform separate analy-
ses at each time-point, using multiple imputation that will incorporate available information from
all evaluation timepoints. As change in FSS from baseline is expected to exhibit substantially
skewed distributions (with many surviving children returning to baseline status), and since under
|
protocol.txt
|
8cd0eafa-ff78-4ed5-8654-f6240aac1d21
|
intention to treat analysis, children who die are treated as having the worst-possible outcome,
rank-based analyses will be used.
6.3.4 Safety Outcomes
The primary safety outcome in the GRACE-2 trial (systemic levels of IL-6 and ferritin) will be
analyzed separately using linear mixed models predicting the laboratory value with a random
|
protocol.txt
|
1b636215-0f1f-48b3-81c8-5d65d226dc21
|
effect for site and subject. Arm, time to laboratory draw, and the interaction between arm
and time will be used to assess whether the GM-CSF arm has a higher rate of laboratory
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkPage 44 of 76 Protocol 90 (Hall, Zuppa and Mourani)
|
protocol.txt
|
c2a0dfe9-82b5-4805-a3de-d154b8c931a5
|
values over time than the placebo arm. The primary safety outcome in the TRIPS trial (rate
of nosocomial infection) will be compared between arms using Poisson regression using
Generalized Estimating Equations with arm as the primary predictor and number of days of
exposure as the offset. Nosocomial infection rates will also be summarized using the pooled
|
protocol.txt
|
fe26ac9d-f8d6-4323-82d6-3c0a1b95d658
|
number of infections per 1000 ICU days. Specific thresholds and/or patterns of systemic
inflammatory biomarker elevation and nosocomial infection rates that would merit a pause in
enrollment or study stoppage will be developed with the DSMB a priori and will be included in
the DSMB charter.
6.3.5 Subgroup Analyses
|
protocol.txt
|
b82a5f59-f7d5-42a9-a77d-5323665617ef
|
Primary, secondary, and exploratory outcome analyses will be repeated by the following sub-
groups to assess possible heterogeneity in treatment effect:
• Age (<1 year, 1 year to onset of puberty, post-pubertal)
• Sex
• Race
• Ethnicity
• Infectious organism type (gram positive bacterium, gram negative bacterium, fungus,
virus, polymicrobial)
• Site of infection
|
protocol.txt
|
19afb422-35bf-43b6-a35c-f5814770a957
|
• Presence of a complex chronic condition
• Presence of baseline immune compromise
• Use of extracorporeal therapies
• Immune subgroup (TRIPS trial)
– Hyperinflammation without immunoparalysis
– Ferritin >2,000ng/ml with immunoparalysis
The outcomes will be re-analyzed using the appropriate statistical methods (e.g., the
|
protocol.txt
|
90670ad3-2418-47ce-abae-87b4090622ad
|
Wilcoxon Rank sum test will be replaced by rank regression for subgroup effect assessment)
and include the main effect of the subgroup and arm and an interaction between the subgroup
and arm. If the interaction p-value is significant at the 0.05 level, this indicates evidence the
effect of arm differs by subgroup level when predicting the outcome.
6.3.6 Handling of Missing Data
|
protocol.txt
|
27a3e896-a5e5-45f1-afa1-63189a0ffbbc
|
Per the intention-to-treat principle, patients who withdraw from the study or are lost to follow-up
will have all available data used in the analysis. Patients whose families withdraw consent to
administer the study agent will be encouraged to allow their child to remain in the trial for
PRECISE Protocol Version 1.07
Protocol Version Date: June 16, 2023
|
protocol.txt
|
25a107ee-9e25-4724-a50b-b9b8a51d6912
|
The Collaborative Pediatric Critical Care Research NetworkProtocol 90 (Hall, Zuppa and Mourani) Page 45 of 76
assessment of outcomes. In the event that a substantial number of patients are withdrawn or
lost to follow-up, baseline characteristics and available information on hospital course and later
follow-up timepoints will be reviewed and compared to patients not withdrawn or lost, to assess
|
protocol.txt
|
cf8cc0c2-eef1-4d84-ba70-78a8853159df
|
empirically whether these patients differ from those remaining in the study for the scheduled
treatment and follow-up time. Missingness for primary, secondary, exploratory, and safety
outcomes will be reviewed in aggregate and by site. Reviews will start as soon as enrollment
opens and will be regulatory monitored so missing data problems can be addressed early in the
study.
|
protocol.txt
|
def1ddb6-0a67-4f4b-b2f5-148a7e838246
|
We expect to observe nearly 100% of outcomes for the primary analysis. PELOD-2 will
be collected daily during the hospital stay, and vital status assessed at 28 days. We expect
minimal loss-to-follow-up for visits after discharge. If substantial missingness is observed,
we will perform multiple imputation to account for missing data in-hospital as well as during
|
protocol.txt
|
4c0f1d40-3e0f-4da7-9173-b6083207a244
|
the follow-up visits. We will minimize loss to follow-up by collecting (and using) multiple
routes of contact with subjects (e.g. email, cell phone, home phone, address); by using outcome
measurement tools that can, together, be completed in 30 minutes; by using local Research
Coordinators to do the follow-up contacts (rather than use a non-local number which families
|
protocol.txt
|
d7ceecc9-98be-4196-be17-211044721752
|
may be reluctant to answer); and by providing financial incentives for each follow-up contact.
7 Data Management
7.1 Clinical Site Data Management
Each clinical site will maintain study records in locked filing cabinets. The site will maintain an
Essential Documents Binder, which may be in paper or electronic form. Copies of all informed
|
protocol.txt
|
40ca05a1-1fb8-466e-8787-3f14b55879ca
|
consent documents will be kept on file and be available for site monitoring inspection (on site or
remote).
7.2 Data Coordinating Center
7.2.1 Data Center Description
The Data Coordinating Center (DCC) in the Department of Pediatrics at the University of Utah
School of Medicine provides data coordination and management services for CPCCRN and
|
protocol.txt
|
611a08fa-1e20-4850-a210-7642054e226d
|
a variety of other national research networks. Anchoring these services is a state-of-the-art,
energy efficient data center. The data center facility supports more than 3500 users around the
country and will provide a secure, reliable, enterprise-wide infrastructure for delivering mission
critical DCC systems and services to CPCCRN.
PRECISE Protocol Version 1.07
|
protocol.txt
|
5ba9a11d-4b19-452e-9d16-5d1eadd5ceff
|
Protocol Version Date: June 16, 2023
The Collaborative Pediatric Critical Care Research NetworkPage 46 of 76 Protocol 90 (Hall, Zuppa and Mourani)
7.2.2 Facility, Hardware, Storage, Data Backup and System Availability
The data center was built using industry standards and energy efficient cooling solutions. The
|
protocol.txt
|
0dfe5a06-6869-4cfe-be7d-f265a8200840
|
data center is cooled by Liquid Cooling Package (LCP) inline cooling technology, providing
the desired efficiency, redundancy and modularity. The LCP utilizes a hot/cold aisle design
that allows for even air distribution to minimize hot spots. The data center’s electrical power
system contains an uninterruptible power supply with a diesel backup generator. The data
|
protocol.txt
|
d6b1f58d-d043-4bfc-82ec-d5be49eb4657
|
center is protected with an FM-200 backed fire suppression system, which provides waterless
fire suppression without leaving behind residue or particulate. Enhanced security measures
are implemented to safeguard the equipment and the data within in it. Security measures are
enforced 24 hours a day, seven days a week, 365 days a year by a combination of on-premise
|
protocol.txt
|
2f54fe80-ea39-4ef7-be74-d02a1336aed8
|
security guards, University police officers, and video surveillance.
The data center has a virtualized environment. This environment consists of more than
415 virtual servers across 25 host servers. Virtualization provides key advantages: (1) High
availability - in the event of hardware failure, virtual machines automatically restart on healthy
|
protocol.txt
|
0a31e8f6-f94b-49b9-9bf0-553ec77ae1b1
|
resources, minimizing impact to end-users; (2) Flexible infrastructure - compute and storage is
seamlessly scaled as current needs change; (3) Rapid deployment - new resources are provi-
sioned on-demand.
Production servers running mission critical applications are clustered and configured for
failover events across multiple clusters. Entire servers are backed-up to a dedicated infrastruc-
|
protocol.txt
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.