Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=iebt20
Expert Opinion on Biological Therapy
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/iebt20
Analytical comparability demonstrated for an
IgG4 molecule, inclacumab, following transfer of
manufacturing responsibility from Roche to Global
Blood Therapeutics
Radu Mihaila, Dipali Ruhela, Lifang Xu, Sandra Joussef, Xin Geng, Jianxia Shi,
Andrea S. Kim, Wendy Yares, Kevin Furstoss & Kent Iverson
To cite this article: Radu Mihaila, Dipali Ruhela, Lifang Xu, Sandra Joussef, Xin Geng, Jianxia
Shi, Andrea S. Kim, Wendy Yares, Kevin Furstoss & Kent Iverson (2022) Analytical comparability
demonstrated for an IgG4 molecule, inclacumab, following transfer of manufacturing responsibility
from Roche to Global Blood Therapeutics, Expert Opinion on Biological Therapy, 22:11, 1417-1428,
DOI: 10.1080/14712598.2022.2143260
To link to this article: https://doi.org/10.1080/14712598.2022.2143260
© 2022 Global Blood Therapeutics.
Published by Informa UK Limited, trading as
Taylor & Francis Group.
View supplementary material
Published online: 09 Nov 2022. Submit your article to this journal
Article views: 153 View related articles
View Crossmark data
ORIGINAL RESEARCH
Analytical comparability demonstrated for an IgG4 molecule, inclacumab, following
transfer of manufacturing responsibility from Roche to Global Blood Therapeutics
Radu Mihaila
a
, Dipali Ruhela
a
, Lifang Xu
a
, Sandra Joussef
a
, Xin Geng
a
, Jianxia Shi
a
, Andrea S. Kim
a
, Wendy Yares
a
,
Kevin Furstoss
a
and Kent Iverson
b
a
Global Blood Therapeutics, South San Francisco, CA, USA;
b
Kent Iverson Consulting, Sonoma, CA, USA
ABSTRACT
Background: Inclacumab is a recombinant, fully human, immunoglobulin IgG4 monoclonal antibody
that selectively binds to P-selectin. Initially discovered and developed by Roche through phase 2 clinical
studies in peripheral arterial disease and coronary artery disease, inclacumab has been in-licensed by
Global Blood Therapeutics (GBT) as a potential treatment to reduce the frequency of vaso-occlusive
crises in individuals with sickle cell disease.
Research design and methods: GBT sought to demonstrate the analytical comparability between
material produced by Roche and material produced by GBT to ensure that no meaningful differences in
identity, safety, purity, potency, or bioavailability exist between the GBT and Roche lots.
Results: Inclacumab samples produced by GBT were found to be comparable to the Roche v0.2
inclacumab samples based on (1) comparable primary and higher-order structures; (2) comparable
purity profiles; (3) comparable potency, in vitro functional activities, and in vivo plasma exposures and
pharmacokinetic profiles; and (4) comparable degradation patterns and kinetics under forced degrada-
tion conditions.
Conclusions: Based on the design of this comparability study and the results obtained, the US Food
and Drug Administration approved the changes to the manufacturing process and gave clearance for
GBT to proceed with phase 3 clinical trials.
ARTICLE HISTORY
Received 29 September
2022
Accepted 31 October 2022
KEYWORDS
Comparability; inclacumab;
monoclonal antibody;
P-selectin; sickle cell disease;
vaso-occlusive crisis
1. Introduction
P-selectin is a cell adhesion molecule expressed on the surface
of platelets and endothelial cells that have been activated in
response to vascular injury and/or inflammation [1,2]. It plays
an essential role in facilitating the multicellular interactions
that occur during thrombotic and inflammatory processes
and is, therefore, implicated in the pathogenesis of numerous
diseases, including coronary artery disease, stroke, diabetes,
and malignancy [2,3]. P-selectin is also known to contribute to
the occurrence of vaso-occlusive crises (VOCs) in patients with
sickle cell disease (SCD) [4,5]. VOCs are thought to be caused
by the adhesion of leukocytes, platelets, and sickled red blood
cells to the endothelium of blood vessels, resulting in vascular
obstruction, tissue ischemia, pain, and in some cases,
death [5,6].
Inclacumab is a recombinant, fully human, immunoglo-
bulin G4 (IgG4) monoclonal antibody (mAb) that selectively
binds to P-selectin, blocking the interaction with its primary
ligand, P-selectin glycoprotein ligand-1 (PSGL-1), and subse-
quently inhibiting P-selectin–mediated adhesive functions
[1,2,7–9]. Proof of concept for the role of P-selectin inhibi-
tion in reducing VOCs in patients with SCD has been estab-
lished by US Food and Drug Administration (FDA) and
European Medicines Agency (EMA) approval of the huma-
nized IgG2 P-selectin mAb crizanlizumab (Adakveo®) [10,11].
Inclacumab has several potential advantages over crizanli-
zumab: a distinct epitope that directly overlaps the PSGL-1
binding site; greater maximal platelet-leukocyte aggregate
inhibition in response to platelet agonists than the same
concentration of crizanlizumab in blood samples from both
healthy volunteers and patients with SCD; and clinical prop-
erties that support a longer dosing interval compared with
crizanlizumab [8–10]. Additionally, two single-point muta-
tions were introduced into the Fc part of inclacumab in
order to avoid antibody-dependent cytotoxicity (L235E)
and to improve structural stability (S228P) [7]. The L235E
point mutation replaces a hydrophobic contact with
a highly charged group, reducing undesirable interactions
by impairing Fc affinity for Fcγ receptors [12,13].
Inclacumab was discovered and developed by Roche; it
was progressed through phase 2 clinical studies in which its
safety and pharmacology was well characterized in more
than 700 individuals, including both healthy volunteers
and patients with cardiovascular disease [14–16]. The pro-
gram was in-licensed from Roche by Global Blood
Therapeutics (GBT) in August 2018 to develop the drug as
a potential treatment to reduce the frequency of VOCs in
individuals with SCD [17,18]. As is typical for mAbs, process
changes in the manufacturing of inclacumab have occurred
throughout its lifecycle [19], resulting in the implementation
CONTACT Radu Mihaila rmihaila@gbt.com Global Blood Therapeutics, 181 Oyster Point Blvd, South San Francisco, CA 94080, USA
Supplemental data for this article can be accessed online at https://doi.org/10.1080/14712598.2022.2143260
EXPERT OPINION ON BIOLOGICAL THERAPY
2022, VOL. 22, NO. 11, 1417–1428
https://doi.org/10.1080/14712598.2022.2143260
© 2022 Global Blood Therapeutics. Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/),
which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.
of three distinct manufacturing processes, referred to here
as Roche version 0.1 (v0.1), Roche v0.2, and GBT. A summary
of the scope of changes introduced among the three differ-
ent processes is given in Table 1.
The Roche v0.1 manufacturing process was used to support
nonclinical and phase 1/2a clinical studies, with some modifi-
cations introduced between the production of nonclinical
Roche v0.1 material and clinical Roche v0.1 material due to
scale-up requirements and process performance improve-
ments. Later, Roche implemented an improved manufacturing
process, designated as Roche v0.2, which offered a higher cell
culture product titer, better purification efficiency, and a drug
substance formulation composition that conferred greater sta-
bility in the final drug product. A comparability assessment of
Roche v0.1 and Roche v0.2 drug substance batches was con-
ducted by Roche to support the implementation of the Roche
v0.2 process. The assessment included routine analyses,
extended characterization of physicochemical properties,
in vitro functional assays, in vivo pharmacokinetic (PK) char-
acterization, and a stress stability study. Roche submitted an
Investigational New Drug application to the FDA for coronary
artery disease in 2012, which included results from the com-
parability assessment. An overview of the scope of the com-
parability assessment is outlined in Table 2. Taken together,
the comparability data demonstrated that the lots produced
with the Roche v0.1 and Roche v0.2 processes were compar-
able, an assessment that was also accepted by the FDA, thus
enabling Roche to implement the v0.2 process to supply
phase 2b clinical trials.
When GBT acquired the rights to develop, manufacture,
and commercialize inclacumab worldwide [20], documenta-
tion regarding the Roche v0.2 manufacturing process and
the analytical procedures and data were transferred to GBT,
along with vials of the Roche v0.2 master cell bank and the
Roche v0.2 reference standard. Because of the technology
transfer, changes made to the v0.2 process to create the GBT
process were minimal. The changes that were introduced were
needed to replace Roche proprietary technologies that were
not transferred to GBT, and the scale of the process was
increased to a size suitable for commercial supply. In addition,
the production of inclacumab was transferred to a new man-
ufacturing site. Process changes were mostly made within the
cell culture and harvest stages of the manufacturing process,
whereas the purification steps are based on the Roche v0.2
process, and the GBT formulation is identical to the v0.2
formulation.
A comprehensive comparability study in alignment with reg-
ulatory guidelines was undertaken to extensively assess the
comparability of inclacumab produced using the Roche and
GBT manufacturing processes, ensuring that no meaningful
differences in identity, safety, purity, potency, and bioavailability
exist between GBT lots to be used in future clinical studies and
the lots used in nonclinical safety and early clinical studies
conducted by Roche. Based on the design of this comparability
study and the results obtained, the FDA approved the changes
to the manufacturing process and gave clearance to proceed
with phase 3 clinical trials in patients with SCD using the incla-
cumab produced by GBT. The approval confirms that the GBT
inclacumab process development, characterization, and analyti-
cal test results are consistent with FDA expectations for a phase
3 mAb therapy. This report presents the results of the aforemen-
tioned comparability assessment between inclacumab pro-
duced by Roche and inclacumab produced by GBT to
demonstrate their comparability and that GBT-produced incla-
cumab is suitable for use in phase 3 clinical trials.
2. Materials and methods
This comprehensive comparability study was executed in
accordance with its protocol, which contained predefined
quality and safety criteria, and the following regulatory guide-
lines: ICH Q5E: Comparability of Biotechnological/Biological
Products Subject to Changes in Their Manufacturing Process,
CPMP/ICH/5721/03 [21]; Guideline on Development,
Production, Characterization and Specifications for Monoclonal
Antibodies and Related Products, EMA/CHMP/BWP/532517/
2008 [22]; and Guideline on Comparability of Biotechnology-
derived Medicinal Products after a Change in the
Manufacturing Process. Non-clinical and Clinical Issues, EMEA/
CHMP/BMWP/101695/2006 [23].
2.1. Inclacumab batches used for comparability
assessment
The materials used for the comparability study consisted of
two batches of drug product from the Roche v0.1 processes
that were produced from independent drug substance
batches, one batch each of drug substance and drug product
(produced from an independent drug substance batch) from
the Roche v0.2 process, and four batches of drug substance
from the GBT process. The four batches of Roche material
were the only historical batches of inclacumab available to
GBT; however, GBT had access to all Roche data (characteriza-
tion, comparability, release, and stability) from all prior Roche
drug substance and drug product lots. These historical data
were leveraged to set the comparability acceptance criteria for
process variability and product stability (degradation pro-
ducts). Roche v0.1 and v0.2 drug products were stored at
2°C to 8°C since the date of manufacture (duration ~9–
12 years), and the Roche v0.2 drug substance was stored
at –70°C since 2011 (duration ~9 years).
2.2. Appearance
For color analysis, inclacumab samples were evaluated visually
under defined viewing conditions and compared to color
references in accordance with European Pharmacopoeia (Ph.
Table 1. Scope of drug substance manufacturing changes across Roche (v0.1
and v0.2) and GBT processes.
Roche v0.1 to Roche v0.2 Roche v0.2 to GBT
Cell line Yes No
Cell line selection system Yes No
Scale of manufacture Yes Yes
Manufacturing site Yes Yes
Purification process Yes No
Formulation Yes No
GBT, Global Blood Therapeutics.
1418 R. MIHAILA ET AL.
Eur.) 2.2.2. For clarity, a quantitative assessment of inclacumab
samples was performed using a calibrated turbidimeter, and
the results were compared to a set of opalescence references
in accordance with Ph. Eur. 2.2.1. Testing was conducted at
Eurofins Lancaster Laboratories, Inc. (Lancaster, PA).
2.3. Osmolality
Osmolality of inclacumab samples was determined via mea-
surement with a calibrated osmometer according to Ph. Eur.
2.2.35 and USP <785>. Testing was conducted at Patheon
(St. Louis, MO).
2.4. Protein content by UV
Total protein concentration of each inclacumab sample was
quantitatively determined via ultraviolet (UV) absorbance
using the SoloVPE system. Absorbance at 280 nm was used
to calculate the concentration of inclacumab using Beer’s law,
A = εlc, where A = absorbance at 280 nm, ε = extinction
Table 2. Outline of the scope of the comparability assessments performed for inclacumab to date.
Type Analytical method
Roche v0.1 vs v0.2
comparability
assessment
Roche v0.1/ v0.2 vs GBT comparability
assessment
Routine analysis Appearance – X
pH – X
Osmolality – X
Protein concentration by UV – X
Peptide map X X
SE-HPLC X –
SE-UPLC – X
CE-SDS (NR) X X
CE-SDS (R) – X
IE-HPLC X –
a
Poloxamer 188 – X
Potency by ELISA – X
Potency by bioassay X X
Bioburden X X
Bacterial endotoxins X X
Sterility X X
HCP content X X
PLBL2 content X –
b
Protein A content X X
DNA content X X
Extended characterization CD, FTIR, DSC – X
SEC-MALS – X
IE-HPLC –
a
X
Sialic acid content – X
ESI-MS analysis X X
LCMS peptide map X X
Peptide mapping (NR) – X
2-AB NP-UHPLC X –
Glycan analysis by fluorescence and MS – X
HCP by 2DGE – X
PLBL2 content –
b
X
Charge isoform characterization Characterization of charge variants between
processes
– X
Thermal stress study (40 °C) Compare degradation pattern X X
In vitro functional characterization Biacore-FcRn binding X X
Octet/FcγR panel – X
Biacore-P-selectin binding X X
Biacore-C1q – X
In vivo PK study in rats Assessment of relative bioavailability X X
In vitro hemolysis testing in human
blood
Effects on blood hemolysis, turbidity, or
precipitation
X –
c
In vivo local tolerability study in
rabbits
Assessment of local tolerability at the injection
sites
X –
c
‘X’ indicates the test was performed; ‘ – ’ indicates the test was not performed.
a
IE-HPLC was performed as part of routine analysis and extended characterization for the comparability test performed by Roche and the comparability test
performed by GBT, respectively.
b
Analysis of PLBL2 content was performed as part of routine analysis and extended characterization for the comparability test performed by Roche and the
comparability test performed by GBT, respectively.
c
Assessment not performed, as there was no change in formulation between the Roche v0.2 process and the GBT process.
2-AB NP-UHPLC, 2-aminobenzamide normal-phase ultra-high-performance liquid chromatography; 2DGE, 2D gel electrophoresis; C1q, complement component 1q;
CD, circular dichroism; CE-SDS, capillary electrophoresis–sodium dodecyl sulfate; DSC, differential scanning calorimetry; ELISA, enzyme-linked immunosorbent
assay; ESI-MS, electrospray ionization–mass spectrometry; FcγR, Fc gamma receptor; FcRn, neonatal Fc receptor; FTIR, Fourier-transform infrared spectroscopy; GBT,
Global Blood Therapeutics; HCP, host cell protein; IE-HPLC, ion-exchange high-performance liquid chromatography; LCMS, liquid chromatography–mass spectro-
metry; MS, mass spectrometry; NR, nonreduced; PK, pharmacokinetic; PLBL2, phospholipase B-like 2; R, reduced; SEC-MALS, size-exclusion chromatography with
multiangle light scattering; SE-HPLC, size-exclusion high-performance liquid chromatography; SE-UPLC, size-exclusion ultra-performance liquid chromatography;
UV, ultraviolet.
EXPERT OPINION ON BIOLOGICAL THERAPY 1419
coefficient of inclacumab (1.54 mL mg
–1
cm
–1
), l = pathlength
(variable), and c = concentration of the sample. In this method,
the absorbance at 320 nm was subtracted from the absor-
bance at 280 nm before calculating the concentration.
A variable pathlength approach utilized the slope (change of
absorbance per change of light pathlength) to calculate the
concentration. The absorbance value was directly proportional
to the protein concentration. Testing was conducted at
Patheon (St. Louis, MO).
2.5. Identity by peptide map
The identity of inclacumab samples was confirmed using gra-
dient reversed-phase high-performance liquid chromatogra-
phy (HPLC). Each sample was reduced, alkylated, digested
with trypsin, and loaded onto an HPLC column. The eluate
was then monitored via UV absorbance. The sample chroma-
togram was compared to the reference material chromato-
gram for presence of any significant new peak(s) or absence
of any characteristic peak(s). A significant peak was defined as
one that is ≥10% of the height of a designated peak as
identified in the test method. Positive identity was confirmed
if the tested material corresponded to the reference material
with respect to retention times and relative intensities of the
15 characteristic peaks chosen as per the test method. Testing
was conducted at Patheon (St. Louis, MO).
2.6. Purity by SE-UPLC
Size-exclusion ultra-performance liquid chromatography (SE-
UPLC) was used to monitor the size heterogeneity of inclacumab
samples under native conditions by employing size-exclusion
chromatography to separate inclacumab aggregates, intact IgG,
and smaller fragments. The assay used an isocratic run with UV
detection at 280 nm, allowing for sensitivity to the analytes
while reducing interference from the buffer components. The
purity of the samples was determined in terms of percentage
area of intact IgG, sum of high-molecular-weight (HMW) forms,
and sum of low-molecular-weight (LMW) forms. Testing was
conducted at Patheon (St. Louis, MO).
2.7. Purity by CE-SDS
Capillary electrophoresis–sodium dodecyl sulfate (CE-SDS) was
used to monitor the size heterogeneity of inclacumab sam-
ples. Samples were heat denatured with sample buffer con-
taining SDS in the presence of beta-mercaptoethanol
(reducing conditions) and in the absence of beta-
mercaptoethanol (nonreducing conditions). Samples (reduced
and nonreduced) were separated by size in a capillary filled
with a polymer matrix in a proprietary buffer that provided
a sieving matrix for separation, causing smaller moieties to
migrate more quickly through the matrix when voltage was
applied. As the molecules moved through the capillary, they
were detected by a photodiode array detector at 220 nm.
Testing was conducted at Patheon (St. Louis, MO).
For the nonreduced sample, percent of IgG was determined
quantitatively in terms of the percent corrected peak area
(CPA) of intact IgG. For the reduced sample, percent of IgG
was determined quantitatively as percent CPA of heavy chain
and light chain.
The CPA calculation for velocity (μV∙s) was performed using
32 Karat software. Peak areas were adjusted for migration
times using the formula LD�A
T¼CPA, where T = migration
time (s), L
D
= capillary length to detector, and A = peak area
(μV∙s). Percent CPA was calculated using the formula %
CPA = (CPA × 100) ÷ (total CPA).
2.8. Poloxamer 188 content by UPLC with charged
aerosol detection
Poloxamer 188 content in inclacumab samples was quanti-
tated using UPLC with a mixed-mode column coupled to
a charged aerosol detector (CAD). The mass-sensitive detector
responded to essentially all nonvolatile and some semi-volatile
compounds eluting from the column. The anion exchange/
reversed-phase column exhibited reduced hydrophobic inter-
actions and allowed the polymeric poloxamer 188 to elute in
a single peak, improving quantitation while also excluding the
overall positively charged protein, thereby minimizing sample
interference. The components of the sample not retained by
the column were sent to waste, utilizing the low flow diverter
on the Corona CAD. The quantitation of poloxamer 188 was
performed by comparing the sample peak area response to
a standard curve. Testing was conducted at Patheon
(St. Louis, MO).
2.9. Potency by binding ELISA
The potency of inclacumab samples was tested using an
enzyme-linked immunosorbent assay (ELISA). Samples were
incubated with biotinylated P-selectin bound to streptavidin-
coated plates, and goat anti-human kappa chain-specific
horseradish peroxidase (HRP)-conjugated detection antibody
was added subsequently. The associated color change was
achieved by addition of the enzyme substrate-chromogen
reagent 3,3ʹ,5,5ʹ-tetramethylbenzidine (TMB). The chromo-
genic signal was proportional to the amount of inclacumab
that bound to P-selectin. Potency was determined by compar-
ing the P-selectin binding of inclacumab relative to the refer-
ence material. Testing was conducted at Patheon
(St. Louis, MO).
2.10. Potency by bioassay
The biological activity of inclacumab samples was tested using
an in vitro leukocyte adhesion bioassay. Varying concentra-
tions of samples and inclacumab reference standard were
incubated with fluorescence-labeled human PSGL-1-expres-
sing cells in a microtiter plate coated with recombinant
human P-selectin. After removal of nonadherent cells, the
remaining adherent cells were lysed, and adhesion was quan-
tified in the form of a fluorescence signal. Potency of the
samples was measured relative to the reference material.
Testing was conducted at Eurofins Lancaster Laboratories,
Inc. (Lancaster, PA).
1420 R. MIHAILA ET AL.
2.11. Host cell protein content
Host cell protein (HCP) content in the inclacumab samples
was measured using a Chinese hamster ovary (CHO) HCP
ELISA kit from Cygnus Technologies. Samples were mixed
with affinity-purified capture anti-CHO antibody immobi-
lized on microtiter strips and an HRP-labeled anti-CHO
detection antibody (goat polyclonal). The resulting reaction
formed a sandwich complex consisting of solid-phase (cap-
ture) antibody, HCP antigen, and enzyme-labeled antibody.
After complex formation, the microtiter strips were washed
to remove any unbound reactants, and TMB, an HRP-
sensitive substrate, was added, with the resulting amount
of hydrolyzed TMB being directly proportional to the
amount of CHO HCP present. The concentration of HCP in
the test sample was calculated from a standard curve gen-
erated using a CHO HCP standard. Testing was conducted at
Patheon (St. Louis, MO).
2.12. Residual protein A content
Residual protein A content in the inclacumab samples was
quantitated using a protein A ELISA kit from RepliGen. Both
inclacumab samples and the protein A standards were
diluted using sample diluent and incubated with immobi-
lized anti-protein antibodies. Protein A ligand was then
detected by the addition of a biotinylated anti-protein
A probe. The high substitution of the probe allowed max-
imum binding of a streptavidin-peroxidase conjugate. The
final detection step involved the use of TMB to give a highly
sensitive colorimetric reaction. The resulting color intensity
was proportional to the amount of protein A ligand present
in the sample. The concentration of protein A in the test
sample was calculated from a standard curve generated
using a protein A standard. Testing was conducted at
Patheon (St. Louis, MO).
2.13. Residual host cell DNA content
Residual host cell DNA content in the inclacumab samples was
determined using a quantitative polymerase chain reaction
(qPCR)-based CHO residual DNA quantitation kit from Life
Technologies. This quantitative assay was based on the real-
time detection of a fluorescence signal during qPCR, and the
amount of signal was directly proportional to the amount of
CHO DNA present. Genomic DNA from in-process and bulk
drug substance samples was recovered using a nucleic acid
extraction kit from Life Technologies. This kit used chemical
lysis and magnetic beads to efficiently extract genomic DNA
from diverse sample types, including samples that contain
high protein and low DNA concentrations. Once DNA was
recovered by elution into a matrix ideal for PCR, it was ampli-
fied via real-time qPCR. The concentration of DNA in the test
sample was calculated by comparing the PCR signal to signals
generated from a standard curve prepared using a CHO geno-
mic DNA standard. Testing was conducted at Patheon
(St. Louis, MO).
2.14. Extended characterization
Inclacumab (Roche v0.1, v0.2, and GBT samples) were assessed
by extended characterization in terms of purity, impurity, and
identity. Size variant distribution was further characterized
using size-exclusion chromatography with multiangle light
scattering (SEC-MALS), and charge heterogeneity was evalu-
ated using ion-exchange HPLC (IE-HPLC). Identity of primary
structure was confirmed using electrospray ionization–mass
spectrometry (ESI-MS) analysis and peptide mapping. Higher-
order structure characterization was performed using Fourier-
transform infrared spectroscopy (FTIR), circular dichroism (CD),
differential scanning calorimetry (DSC), and disulfide mapping.
Post-translational modifications were assessed using peptide
map liquid chromatography–mass spectrometry (LCMS),
released glycan profiles were analyzed via hydrophilic-
interaction liquid chromatography (HILIC)-MS, and sialic acid
content was evaluated using UPLC. Testing was conducted at
KBI Biopharma, Inc. (Louisville, CO) and Patheon
(St. Louis, MO).
2.15. In vitro functional characterization
To assess antigen binding, inclacumab (Roche v0.1, v0.2, and
GBT samples) was captured on Biacore chips via anti–human
Fc antibodies attached to the chip surface using covalent
amine-coupling chemistry. The antigen, P-selectin, was then
injected over the chip at various concentrations, and the
binding response was measured in terms of resonance units
(RU). The contact time (in seconds) with the chip was opti-
mized for each individual inclacumab sample to ensure that
the level of inclacumab captured on the chip was comparable
across the samples. Binding response difference was measured
by comparing the maximum response of each sample. The
formula used to calculate the RU maximum difference is as
follows: relative RU max difference = [RU max (sample)/RU
max (reference standard)] × 100%. Testing was conducted at
Sartorius Stedim North America Inc. (Cambridge, MA). To
assess neonatal Fc receptor (FcRn) binding, FcRn was directly
immobilized onto the Biacore chip surface using covalent
amine-coupling chemistry. Test article containing inclacumab
(Roche v0.1, v0.2, and GBT samples) was injected over the
chip, and binding was measured in RU. Tysabri (natalizumab)
was used as the IgG4 control, and a GBT sample lot served as
the internal reference standard. The formula used to calculate
relative FcRn binding (in terms of RU maximum difference) is
as follows: relative RU max difference = [RU max (sample)/RU
max (reference standard)] × 100%. Testing was conducted at
Sartorius Stedim North America Inc. (Cambridge, MA).
To assess Fc gamma receptor (FcγR) binding, Fc receptors
were immobilized on a biosensor, and the association/disso-
ciation rates (k
off
/k
on
) of inclacumab samples (Roche v0.1, v0.2,
and GBT samples) were measured to determine the binding
affinity, which was calculated using an appropriate curve-
fitting model. An in-house IgG1 control was confirmed to
bind with FcγRI, FcγRIIA, FcγRIIB, FcγRIIIA-F, FcγRIIIA-V, and
FcγRIIIB. Testing was conducted at LakePharma, Inc. (San
Carlos, CA).
EXPERT OPINION ON BIOLOGICAL THERAPY 1421
To assess C1q binding, inclacumab (Roche v0.1, v0.2, and
GBT samples) was immobilized on the surface of a Series
S CM5 Sensor Chip using amine-coupling chemistry and
a defined immobilization method. Rituximab (IgG1 isotype)
and nivolumab (IgG4 isotype) were also immobilized and
served as appropriate positive and negative controls for C1q
binding, respectively. C1q was flowed across the biosensor
surface to permit binding at various concentrations to deter-
mine k
on
over time; k
off
was determined upon ceasing the
flow. When quantifiable binding was observed, the affinity of
C1q to the sample (equilibrium dissociation constant, K
D
) was
calculated from k
off
and k
on
. Testing was conducted at
Sartorius Stedim North America Inc. (Cambridge, MA).
2.16. In vivo characterization
Inclacumab (Roche v0.1, v0.2, or GBT process batch) was
administered as a single intravenous bolus dose of 30 mg/
kg to male Wistar rats (n = 12/group). Blood samples for PK
evaluation were collected from all animals predose and at 1,
7, 24, 48, 96, 168, 336, 432, 528, and 672 hours postdose.
Blood was maintained in chilled cryoracks and centrifuged
within 1 hour to obtain plasma. Collected plasma was
placed on dry ice prior to storage at approximately – 70°C.
Plasma concentrations of inclacumab were quantified using
a validated Meso Scale Discovery electrochemiluminescence
immunoassay with a lower limit of quantification of 205 ng/
mL. The PK parameters for each group were calculated via
noncompartmental analysis using the intravascular input
model and linear-up log-down area under the concentra-
tion–time curve (AUC) calculation method in Phoenix version
64 (Certara USA, Inc., Princeton, NJ). Plasma sample concentra-
tion results below the lower limit of quantitation were
reported as such. These values were treated as 0 in PK
calculations.
Equivalence analysis was conducted to assess bioequiva-
lence of the GBT material and the Roche v0.2 material in
terms of maximum concentration (C
max
) and AUC of incla-
cumab exposures in plasma (area under the concentration–
time profile from time 0 to the last quantifiable concentra-
tion [AUC
last
] and area under the concentration–time profile
from time 0 to infinite time [AUC
0-∞
]). A comparison of
Roche v0.1 material to Roche v0.2 material was also per-
formed. To meet acceptance criteria for material compar-
ability, the 90% confidence intervals (CIs) for the ratio of the
geometric means for C
max
and AUC were required to fall
within the standard bioequivalence limits of 80% to 125%
[24]. Inferences assumed log-normal distribution of data.
Statistical analysis was implemented via PROC TTEST in
SAS v9.4.
In vivo testing was conducted at Covance Laboratories Inc.
(Madison, WI) in accordance with their applicable lab standard
operating procedures. Plasma sample analysis was conducted
at PPD Laboratories (Richmond, VA).
2.17. Thermal stress testing
Stress stability testing of inclacumab samples was performed
at 40°C and 75% relative humidity using 2 Roche v0.2 samples
and 2 GBT samples. GBT samples were assessed at weeks 0, 2,
4, 6, 8, 12, and 24. Due to limited availability of Roche v0.2
material, Roche v0.2 samples were only tested at weeks 0, 4,
12, and 24. Samples were monitored using 8 test methods:
appearance, SE-UPLC, IE-HPLC, CE-SDS (nonreduced/reduced),
a cell-based potency assay, protein concentration, and polox-
amer 188 concentration. Testing was conducted at Patheon
(St. Louis, MO) and Eurofins Lancaster Laboratories, Inc.
(Lancaster, PA).
3. Results
3.1. Comparability assessment plan
While the changes that occurred between the v0.1 and v0.2
processes were more extensive compared with those between
the v0.2 and GBT processes, the comparability assessment
supporting the transition from the v0.2 to GBT process was
designed to be more comprehensive, reflecting a scope
appropriate for a product advancing to phase 3 studies [19]
(Table 2). The comparability assessment included inclacumab
produced using the GBT process and the Roche v0.1 and v0.2
processes. The comparability assessment was primarily
designed to compare GBT batches to Roche v0.2 batches, as
the GBT manufacturing process is based on the Roche v0.2
process. Roche v0.1 materials were also included in the com-
parability assessment, given GBT’s intention to leverage those
data to progress to a phase 3 study. Comparability was
assessed in terms of general properties, purity and impurity
profiles, identity, potency, binding activity, bioavailability, and
kinetics of degradation using data from routine release test-
ing, extended characterization, charge isoform characteriza-
tion, in vitro functional testing, an in vivo PK study, and
thermal stress testing. Acceptance criteria for comparability
with respect to quality and safety were predefined.
3.2. Routine release testing for comparability
assessment
Routine test methods used for the comparability assessment
and a summary of results from those tests are shown in
Table 3. Key assessments are described in more detail in the
following sections.
3.2.1. General properties
The composition and strength of drug products and sub-
stances were considered obligatory critical quality attributes
(CQAs) included in routine release testing. Appearance, pH,
osmolality, protein concentration, and poloxamer 188 con-
centration were comparable for Roche v0.2 and GBT materi-
als. All samples met the set acceptance criteria or the
specification set for their respective formulation
compositions.
3.2.2. Purity and impurities
3.2.2.1. Size variants.
HMW species and LMW species can affect potency and
immunogenicity and are thus considered CQAs. The mole-
cular size distribution of the inclacumab samples was
1422 R. MIHAILA ET AL.
assessed using SE-UPLC and capillary electrophoresis–
sodium dodecyl sulfate (CE-SDS) analyses, which showed
that the main component was the intact IgG. Small changes
(<1.2%) in HMW species and LMW species were observed
upon long-term storage under non-frozen conditions of
both v0.1 samples and one of the v0.2 samples. The
Roche v0.2 sample stored at – 70°C and GBT samples had
comparable levels of HMW species and LMW species.
3.2.2.2. Process-related impurities. Process-related impuri-
ties encompass those that are derived from cell substrates
(eg, HCPs, host cell DNA), cell culture (eg, selection agents
and other media components), and downstream processing
(eg, leached protein A ligand), and are considered CQAs
[25]. HCPs, residual protein A, and residual DNA are of
particular concern, as they are the most commonly encoun-
tered process-related impurities, and they can impact the
safety and immunogenicity of an antibody [19]. Assessment
of residual HCP content showed fewer and no new HCPs in
the GBT samples compared with the Roche samples,
demonstrating a more robust clearance of HCPs, attributa-
ble to improvements made in the GBT purification process.
All GBT and Roche inclacumab samples had residual DNA
and protein A below the limit of quantitation of the respec-
tive analytical methods.
3.2.3. Potency
Performed as part of routine testing, a cell-based potency
assay, exploiting the ability of inclacumab to inhibit binding
of human promyeloblast cells (expressing PSGL-1) to immo-
bilized P-selectin, and a biotinylated P-selectin-based anti-
gen-binding ELISA demonstrated that all inclacumab
samples produced by GBT were equipotent. The cell-based
assay was transferred to GBT from Roche when inclacumab
was in-licensed. The ELISA method is more precise and has
a tighter acceptance criterion than the cell-based assay. All
samples met the set acceptance criteria; however, the rela-
tive potencies of the Roche v0.1 samples were lower than
those reported by Roche in the original release data. The
decrease in potency of the Roche v0.1 samples correlated
with the higher oxidation level in the L8 peptide within the
antigen-binding region (measured by peptide mapping with
LCMS); the higher oxidation level may be attributable to
peroxides generated during the extended non-frozen
storage.
3.3. Extended characterization for comparability
assessment
The extended characterization methods used for the compar-
ability assessment and a summary of results from those tests
Table 3. Routine testing summary for comparability.
Analytical
method Assessment of Summary of results
a
Appearance Color and clarity All samples were comparable and met the acceptance criteria
pH pH All samples met the specification set for their respective
formulation compositions; Roche v0.2 and GBT samples were
comparable
Osmolality Osmolality All samples were comparable and met the acceptance criteria
Protein
concentration
by UV
Concentration of protein All samples met the specification set for their respective
formulation compositions; Roche v0.2 and GBT samples were
comparable
Peptide map Identity/peptide pattern All samples were confirmed with positive identity; samples had
comparable UV profiles
SE-UPLC Intact IgG, LMW, and HMW species All samples were comparable and met the acceptance criteria
CE-SDS (NR/R) Size variants (nonreducing conditions); light chain/heavy chain proportions
(reducing conditions)
All samples were comparable and met the acceptance criteria
Roche v0.1samples, which were stored under non-frozen
conditions for ~9–12 years, showed a slight decrease in
purity
cIEF Charge heterogeneity caused by deamidation and/or N-/C-terminal
heterogeneity of the heavy chain
All samples met the acceptance criteria
GBT samples had a slightly higher percentage of basic
species compared to Roche v0.1 and v0.2 samples
All samples had the same main peak pI
Poloxamer 188 Excipient content Roche v0.2 and GBT samples were comparable and met the
acceptance criteria
Not applicable to Roche v0.1 samples, as they did not contain
poloxamer 188
Potency by ELISA Binding to P-selectin, depicting the mechanism of action of inclacumab All samples were comparable and met the acceptance criteria
Potency by
bioassay
Relative potency with respect to inhibition of binding of human
promyeloblast cells (expressing P-selectin glycoprotein ligand-1) to
immobilized P-selectin
All samples met the acceptance criteria
All samples were comparable except for one Roche v0.1
sample, which trended lower
HCP HCP removal during purification Reduced levels of HCP in the GBT samples compared to Roche
v0.1 and v0.2 samples. GBT samples were below LOQ
Residual protein
A
Protein A removal during purification GBT and Roche samples were below LOQ
Residual host
cell DNA
DNA removal during purification GBT and Roche v0.2 samples were below LOQ
a
The acceptance criteria for comparability were same as the batch release drug substance specifications.
CE-SDS (NR/R), capillary electrophoresis–sodium dodecyl sulfate (nonreduced/reduced); cIEF, capillary isoelectric focusing; ELISA, enzyme-linked immunosorbent
assay; GBT, Global Blood Therapeutics; HCP, host cell protein; HMW, high molecular weight; IgG, immunoglobulin G; LMW, low molecular weight; LOQ, limit of
quantitation; pI, isoelectric point; SE-UPLC, size-exclusion ultra-performance liquid chromatography; UV, ultraviolet.
EXPERT OPINION ON BIOLOGICAL THERAPY 1423
are shown in Table 4. Key assessments are described in more
detail in the following sections.
3.3.1. Purity and impurities
3.3.1.1. Size variants. SEC-MALS analysis was performed to
confirm the accuracy of the UV detection used in the routine SE-
UPLC method and to characterize the size variant distribution of
Roche v0.1, Roche v0.2, and GBT samples. All samples were
shown to have comparable intact IgG molecular weights, and
the levels of aggregate present in all of the samples were very
low and comparable to the levels detected using the UV method.
3.3.1.2. Charge heterogeneity.
Like other mAbs, inclacumab is heterogeneous and contains
product-related variants that manifest during the manufactur-
ing process [19]. A range of product isoforms is created by
varying proportions of post-translational modifications that
occur as part of product expression during cell culture produc-
tion and due to stability and product recovery effects during
harvest, purification, formulation, and storage [19]. Some of
these modifications alter the charge distribution on the sur-
face of mAbs and result in isoforms with different net surface
charges. The net surface charge is used to define isoforms as
either acidic, basic, or the main species [19]. The post-
translational modifications that create differences in net sur-
face charge can potentially affect the biological activity of an
antibody, making the charge isoform profile an important
feature to characterize with respect to product quality and
comparability [19]. Furthermore, the distribution of charge
isoforms can be sensitive to process changes introduced dur-
ing manufacturing [19]. Based on these considerations,
a charge isoform profile is typically designed to be a CQA.
Table 4. Extended characterization summary for comparability.
Analytical
method Assessment of Acceptance criteria Summary of results
CD, FTIR, DSC Secondary and tertiary structures Comparable spectra within
experimental error
All samples had comparable spectra and displayed
similar unfolding events
SEC-MALS Intact IgG and aggregates Light scattering and UV show
comparable amounts of
aggregates
All samples had comparable intact IgG molecular weight
IE-HPLC Charge heterogeneity caused by post-
translational modifications
Main peak (≥55.0%)
Acidic region (≤40.0%)
Basic region (≤30.0%)
(Based on Roche IE-HPLC
release specification)
All samples met specification except for one Roche v0.1
sample, which may be due to storage under non-
frozen conditions for an extended period of time and
lower stability of the polysorbate 20 formulation
GBT samples had slightly higher percentage of basic
species compared to Roche v0.1 and v0.2 samples
Sialic acid
content
NANA and NGNA levels Report NANA and NGNA levels All samples had nondetectable NGNA levels
GBT samples had NANA levels that fell between
Roche v0.2 and Roche v0.1
Electrospray
ionization–
mass
spectrometry
analysis
Molecular masses predicted for the intact
molecule, the light chain, and the heavy chain
Report All samples had comparable intact molecular masses
except for one Roche v0.1 sample due to storage
under non-frozen conditions for an extended period
of time and lower stability of the polysorbate 20
formulation
LCMS peptide
map
Primary structure; microheterogeneity due to
post-translational modifications (including
deamidation, oxidation, and N-/C-terminal
variants)
Report All samples have comparable PTMs; exceptions include:
Roche v0.1 samples had higher oxidation levels and
slightly higher deamidation levels compared with
other samples due to storage under non-frozen
conditions for an extended period of time, and lower
stability of the polysorbate 20 formulation
Roche v0.1 and Roche v0.2 had higher levels of
C-terminal lysine clipping compared with GBT
samples
Nonreduced
peptide
mapping
Locations of disulfide linkages Report All samples had identical disulfide linkages
Glycan analysis
by
fluorescence
and MS
Oligosaccharide pattern No new species in the GBT
samples above 5% compared
to Roche v0.1 and v0.2
samples
Same (no new) glycan species observed in GBT samples
compared with Roche samples
GBT samples had galactosylation levels that fell
between Roche v0.1 and Roche v0.2 levels
Comparable fucosylation levels observed in all
samples
HCP by 2DGE HCP 2DGE pattern Report No new and fewer HCPs observed in GBT samples
compared with Roche samples
PLBL2 (HCP) by
ELISA
PLBL2 quantitation ≤50 ng/mg Reduced levels of PLBL2 in the GBT samples compared
to Roche v0.1 and v0.2 samples. GBT samples had
levels below LOQ
2DGE, 2D gel electrophoresis; CD, circular dichroism; DSC, differential scanning calorimetry; ELISA, enzyme-linked immunosorbent assay; FTIR, Fourier-transform
infrared spectroscopy; GBT, Global Blood Therapeutics; HCP, host cell protein; IE-HPLC, ion-exchange high-performance liquid chromatography; IgG, immunoglo-
bulin G; LCMS, liquid chromatography–mass spectrometry; LOQ, limit of quantitation; MS, mass spectrometry; NANA, N-acetylneuraminic acid; NGNA,
N-glycolylneuraminic acid; PLBL2, phospholipase B-like 2; PTM, post-translational modification; SEC-MALS, size-exclusion chromatography with multiangle light
scattering; UV, ultraviolet.
1424 R. MIHAILA ET AL.
The charge isoform profiles of the inclacumab compar-
ability assessment samples were evaluated using IE-HPLC. All
inclacumab samples produced by GBT had a consistent iso-
form profile, with highly similar peak percentage values. The
GBT samples contained no new detectable charge isoforms
compared with the Roche samples, though some differences
in the proportions of charge isoforms were observed. The
historical v0.1 values from original Roche testing showed
higher percentages of acidic species than GBT and Roche
v0.2 lots, which further increased upon storage under non-
frozen conditions. No such change in percentage of acidic
species was observed for Roche v0.2 lots, consistent with the
expected improvement in the stability of the Roche v0.2
formulation. GBT samples had slightly higher levels of basic
species. This may be attributable to a lower proportion of
C-terminal lysine clipping of the heavy chain (see the post-
translational modification analysis section for further detail).
3.3.2. Identity
Characterization of primary structure along with higher-order
structure and post-translational modifications was performed
as part of the extended characterization testing.
3.3.2.1. Primary structure. The expected primary structure
of inclacumab was confirmed via ESI-MS analysis of intact and
reduced inclacumab and via LCMS analysis that employed
a trypsin/Lys-C/Glu-C digestion. MS analysis confirmed that
the molecular masses of all samples analyzed were consistent
with predicted masses deduced from the amino acid sequence
of inclacumab, except for 1 Roche v0.1 sample that had
a greater mass than predicted (which may have been due to
the oxidation of the material under long-term nonfrozen sto-
rage). LCMS trypsin/Lys-C/Glu-C peptide mapping confirmed
the amino acid sequence for inclacumab with 100% sequence
coverage.
3.3.2.2. Higher-order structure. The higher-order structure
of inclacumab in the comparability samples was characterized
using FTIR and CD to determine the secondary and tertiary
structures of the molecule, respectively. DSC was applied to
assess the thermal and conformational stability of inclacumab
in the comparability samples. Results from these tests showed
that all samples had highly similar structures and thermal
stability. Disulfide mapping also confirmed that Roche and
GBT materials share common patterns of disulfide linkages
and unpaired cysteines.
3.3.2.3. Post-translational modifications. Post-translational
modifications (PTMs), including methionine oxidation, deami-
dation of asparagine residues, cyclization of glutamine to form
pyroglutamate at the N terminus of the heavy chain, and
intact and absent C-terminal lysine, were detected and quan-
tified using peptide mapping analysis with LCMS. Inclacumab
present in Roche v0.1, Roche v0.2, and GBT samples had the
same types of PTMs at the same sequence locations. The
proportions of modifications were comparable across the sam-
ples, except increased C-terminal lysine clipping was observed
in both Roche v0.1 and v0.2 samples compared with the GBT
samples and, while GBT samples contained the same glycan
species as Roche v0.1 and v0.2 samples, its glycan profile was
between those of the Roche v0.1 and v0.2 samples. Results
characterizing the glycan profile of inclacumab samples using
a HILIC/fluorescence method coupled with MS detection were
generally consistent with the LCMS PTM characterization. This
method showed that the relative proportions of glycans were
different across different inclacumab sources, with Roche v0.1
samples having the highest levels of galactosylated glycans,
the Roche v0.2 samples having the lowest levels of galactosy-
lated glycans, and GBT samples having intermediate levels of
galactosylated glycans. Results were consistent across samples
from the same source (ie, Roche v0.1, Roche v0.2, or GBT).
These differences detected in C-terminal lysine clipping and
glycan profiles did not have any effect on the functional
activity of GBT samples, which was comparable to that of
Roche v0.2 samples based on in vitro and in vivo
characterization.
Sialylation is an important characteristic of glycoproteins
that can affect their potency, half-life, and immunogenicity by
salvaging glycoproteins from degradation [26]. In general, all
inclacumab samples contained a low abundance of sialic acid
(ie, N-acetylneuraminic acid and N-glycolylneuraminic acid),
with low levels assumed to be functionally insignificant in
eliciting an immune response [27].
3.4. In vitro and in vivo nonclinical testing for
comparability assessment
The scope and results of the studies used to compare in vitro
biological functions and in vivo PK are shown in Table 5.
Overall, the results of the in vitro and in vivo studies demon-
strated comparable functional activities, plasma exposures,
and PK profiles across GBT and Roche v0.2 samples. Key
observations from the comparability assessments are
described in more detail in the following sections.
3.4.1. In vitro functional characterization
Surface plasmon resonance technology (Biacore and Octet) was
used to assess antigen and effector function receptor binding
affinity. All GBT samples produced similar P-selectin, FcRn, and
FcγR binding affinities, and the binding affinities across GBT and
Roche v0.2 samples were comparable despite differences
observed in the distribution of post-translation isoforms and
charge variants. GBT and Roche v0.1 and v0.2 samples had com-
parable binding to the FcγR panel, though the binding affinity
levels for the inclacumab samples were lower than those for
a commercial IgG4 molecule (nivolumab) containing the L235E
mutation. GBT and Roche samples are highly fucosylated (above
89%), which may further impair binding to FcγRs compared to
fucose-deficient antibodies [28]. All inclacumab samples did not
bind to complement component 1q, as expected.
3.4.2. In vivo characterization
To evaluate any potential effects of the differences observed
during the analytical comparability on the relative bioavail-
ability of inclacumab in vivo, a PK study in rats was performed.
The mean inclacumab plasma concentration–time profiles in
male rats after a single intravenous dose of inclacumab from
EXPERT OPINION ON BIOLOGICAL THERAPY 1425
GBT, Roche v0.1, or Roche v0.2 production processes were
closely parallel to each other (Figure 1). All three manufactur-
ing lots of inclacumab showed similar plasma PK parameter
values (Supplemental Table 1). No substantial differences
were observed in plasma exposures (90% CIs for C
max
,
AUC
last
, and AUC
0-∞
ratios were within the bioequivalence
acceptance criteria limits of 80% to 125%) or overall PK pro-
files for inclacumab derived from the GBT and the Roche v0.2
production processes, demonstrating that the two processes
are comparable in this setting. Acceptance criteria for
bioequivalence were also met for the comparison of Roche
v0.1 material to Roche v0.2 material.
3.5. Thermal stress testing for comparability assessment
Roche v0.2 samples and GBT samples were compared side-by-
side in a stress stability study conducted at 40°C and 75%
relative humidity. Samples of Roche v0.1 material were not
included in this study due to the formulation composition
difference and the higher oxidation level observed in the
Table 5. In vitro and in vivo testing for comparability.
Type
Analytical
method Assessment of Acceptance criteria Summary of results
In vitro functional
characterization
Biacore-FcRn
binding
Binding to FcRn (a
key factor for
in vivo half-life)
Comparable binding to FcRn within
the same order of magnitude
All Roche v0.2 and GBT samples had comparable k
on
, k
off
, and
KD to FcRn
Roche v0.1 samples had lower binding to P-selectin
compared with Roche v0.2 and GBT samples due to
degradation of v0.1 samples during long-term non-frozen
storage
Octet/FcγR panel Binding to FcγRs Comparable binding to FcγR panel
within the same order of
magnitude
All samples had comparable KD to FcγRI and FcγRIIB
No detectable binding was observed to FcγRIIA, FcγRIIIA-F,
FcγRIIIA-V, or FcγRIIIB
Biacore
P-selectin
binding
Binding to
P-selectin
Relative binding (80%-120%) All Roche v0.2 and GBT samples had comparable k
on
, k
off
, and
KD to P-selectin
Roche v0.1 samples showed ~10%-20% lower binding to
P-selectin compared with Roche v0.2 and GBT samples,
attributed to degradation of v0.1 samples during long-term
non-frozen storage
Biacore C1q
binding
Binding to C1q No binding No binding was observed for Roche v0.1, Roche v0.2, and GBT
samples
In vivo
characterization
Pharmacokinetic
profile in rats
Assessment of
relative
bioavailability
in vivo
90% CIs for the ratio of the
geometric means for C
max
and
AUC to be within 80%-125%
GBT and Roche materials met the acceptance criteria
AUC, area under the concentration–time curve; C1q, complement component 1q; CI, confidence interval; C
max
, maximum plasma concentration; FcγR, Fc gamma
receptor; FcRn, neonatal Fc receptor; GBT, Global Blood Therapeutics; KD, equilibrium dissociation constant; k
off
, dissociation rate constant; k
on
, association rate
constant.
Figure 1. Male rats were administered a single intravenous dose of inclacumab from GBT, Roche v0.1, or Roche v0.2 production processes (30 mg/kg). The plasma
concentration of inclacumab in micrograms per milliliter on a logarithmic scale is plotted versus time in hours after dosing. Data points show the mean plasma
concentration from 12 samples per group, with error bars representing standard deviation. The lines for all three process materials are closely parallel to each other,
decreasing from approximately 600 to 700 μg/mL 1 hour after the dose to approximately 70 to 80 μg/mL 672 hours the after dose. Comparisons of GBT material to
Roche v0.2 material and Roche v0.2 material to Roche v0.1 material resulted in 90% CIs of the C
max
, AUC
last
, and AUC
0-∞
ratios within the bioequivalence limits of
80% to 125%. AUC
0–∞
, area under the concentration–time curve from time 0 to infinity; AUC
last
, area under the concentration–time curve from time 0 to the time at
which the last quantifiable concentration was observed; C
max
, maximum plasma concentration; CI, confidence interval; GBT, Global Blood Therapeutics; SD, standard
deviation.
1426 R. MIHAILA ET AL.
available Roche v0.1 materials. This study was performed to
compare the stability and degradation behavior of the sam-
ples produced with Roche v0.2 and GBT manufacturing pro-
cesses after storage for 24 weeks at 40°C. Results of the
thermal stress data demonstrated that GBT and Roche v0.2
materials had comparable degradation pathways, rates, and
end products, as determined by stability-indicating methods
(SE-UPLC, IE-HPLC, CE-SDS [nonreduced/reduced], and a cell-
based potency assay).
4. Discussion
Inclacumab samples produced by GBT were found to be
comparable to the Roche v0.2 inclacumab samples based
on (1) comparable primary and higher-order structures; (2)
comparable purity profiles; (3) comparable potency,
in vitro functional activities and in vivo plasma exposures
and PK profiles; and (4) comparable degradation patterns
and kinetics under forced degradation conditions. The
study detected several differences between the materials
produced by the different manufacturing processes,
though the totality of data confirmed that these differ-
ences had no effect on product quality or potency of
inclacumab produced by GBT and that it is suitable for
its intended use in phase 3 clinical trials.
Differences in charge distribution were detected, with
GBT material having a lower proportion of acidic charge
isoforms than Roche v0.1 material but a comparable pro-
portion to Roche v0.2 material. GBT performed additional
charge variant characterization on batches from all three
processes for a deeper comparability assessment of charge
isoform composition of Roche and GBT materials. The
purpose of the additional charge isoform characterization
study was to increase the detectability of differences in
isoform composition across the samples and to develop
an initial understanding of the potential functional impli-
cations of the charge isoform profile changes; however,
results obtained for the charge variant studies are beyond
the scope of this publication.
GBT material also had a slightly higher proportion of
basic species due to reduced clipping of C-terminal lysine;
however, these differences did not affect biological activ-
ity. Glycan distribution levels and sialic acid in the form of
N-acetylneuraminic acid levels for GBT materials fell
between those of Roche v0.1 and v0.2 materials, but no
new glycoforms were observed, and differences did not
affect activity. Roche v0.1 materials stored for more than
10 years at 2°C to 8°C showed oxidation in the L8 (Trp
94
)
peptide that may affect P-selectin binding and oxidation
in the H23 (Met
256
) peptide that may affect FcRn binding.
In addition to demonstrating comparability, this study
also contributes to the scientific understanding of quality
attributes of inclacumab, as well as of IgG4 antibodies in
general. Results from the GBT comparability assessment
were submitted in the GBT Investigational New Drug
Application for inclacumab, which was approved; GBT is
currently enrolling patients with SCD in phase 3 clinical
studies of inclacumab [17,18].
5. Conclusions
In summary, GBT designed an extensive analytical comparabil-
ity plan, in alignment with the regulatory guidelines that
rigorously characterized both Roche and GBT batches of incla-
cumab to demonstrate consistency between the GBT lots and
the lots used in preclinical and clinical studies by Roche.
Inclacumab samples produced by GBT were found to be com-
parable to the Roche v0.2 inclacumab samples across all ana-
lytical methods used for testing. Based on these results, the
FDA approved the changes to the manufacturing process and
gave clearance for GBT to proceed with phase 3 clinical trials
of inclacumab in patients with SCD.
Acknowledgments
We thank Nadine Ritter, PhD, for reviewing the manuscript. Medical writ-
ing and editorial assistance were provided by Lindsay Tannenholz, PhD
(Healthcare Consultancy Group), funded by Global Blood Therapeutics.
Funding
This paper was funded by Global Blood Therapeutics, a wholly owned
subsidiary of Pfizer Inc.
Declaration of interest
R Mihaila, D Ruhela, L Xu, S Joussef, X Geng, AS Kim, W Yares, and K Furstoss
are all employees of Global Blood Therapeutics. J Shi is a former employee of
Global Blood Therapeutics. The current affiliation for J Shi is Ideaya
Biosciences, South San Francisco, CA, USA. K Iverson is a consultant for
Global Blood Therapeutics. Global Blood Therapeutics is a wholly owned
subsidiary of Pfizer Inc. The authors have no other relevant affiliations or
financial involvement with any organization or entity with a financial interest
in or financial conflict with the subject matter or materials discussed in the
manuscript apart from those disclosed.
Reviewer disclosures
Peer reviewers on this manuscript have received an honorarium from
Expert Opinion on Biological Therapy for their review work, but have no
other relevant financial relationships to disclose
Author contributions
Conceptualization: K Furstoss, K Iverson; methodology: R Mihaila, X Geng,
J Shi; investigation: R Mihaila, L Xu, S Joussef, Xin G, J Shi; data curation:
R Mihaila, D Ruhela, L Xu; visualization: R Mihaila, D Ruhela; writing –
original draft preparation: R Mihaila, D Ruhela, K Iverson; writing – review
and editing: all authors. All named authors meet the International
Committee of Medical Journal Editors criteria for authorship for this article,
take responsibility for the integrity of the work as a whole, and have given
their approval for this version to be published.
Data availability
The authors confirm that the data supporting the findings of this study are
available within the article.
References
Papers of special note have been highlighted as either of interest (•) or of
considerable interest (••) to readers.
EXPERT OPINION ON BIOLOGICAL THERAPY 1427
1. Kutlar A, Embury SH. Cellular adhesion and the endothelium:
p-selectin. Hematol Oncol Clin North Am. 2014;28(2):323–339.
2. André P. P-selectin in haemostasis. Br J Haematol. 2004;126
(3):298–306.
3. Kappelmayer J, Nagy B Jr., Miszti-Blasius K, et al. The emerging
value of P-selectin as a disease marker. Clin Chem Lab Med.
2004;42(5):475–486.
4. Ataga KI, Kutlar A, Kanter J, et al. Crizanlizumab for the prevention
of pain crises in sickle cell disease. N Engl J Med. 2017;376
(5):429–439.
5. Kato GJ, Piel FB, Reid CD, et al. Sickle cell disease. Nat Rev Dis
Primers. 2018;4:18010.
6. Rees DC, Williams TN, Gladwin MT. Sickle-cell disease. Lancet.
2010;376(9757):2018–2031.
7. Schmitt C, Abt M, Ciorciaro C, et al. First-in-man study with incla-
cumab, a human monoclonal antibody against P-selectin.
J Cardiovasc Pharmacol. 2015;65(6):611–619.
8. Geng X, Mihaila R, Yuan Y, et al. Inclacumab, a fully human
anti-P-selectin antibody, directly binds to PSGL-1 binding region
and demonstrates robust and durable inhibition of cell adhesion.
Presented at: American Society of Hematology Annual Meeting &
Exposition; 2020 Dec 5–8; virtual. Poster 1707.
9. Mayer C, Cooper DS, Redfern A, et al. Preliminary results of a phase
1 study in healthy subjects administered inclacumab, a fully human
IgG4 anti-P-selectin monoclonal antibody in development for treat-
ment of sickle cell disease. Presented at: American Society of
Hematology Annual Meeting & Exposition; 2021 Dec 11; Atlanta,
GA, USA, and virtual. Poster 977.
10. Adakveo. Prescribing information. Novartis Pharmaceuticals
Corporation. 2021.
11. Adakveo. Summary of product characteristics. Novartis
Pharmaceuticals Corporation. 2020.
12. Hanson QM, Barb AW. A perspective on the structure and receptor
binding properties of immunoglobulin G Fc. Biochemistry. 2015;54
(19):2931–2942.
13. Oganesyan V, Gao C, Shirinian L, et al. Structural characterization of
a human Fc fragment engineered for lack of effector functions. Acta
Crystallogr D Biol Crystallogr. 2008;64(Pt 6):700–704.
14. GBT provides regulatory and pipeline updates in sickle cell disease
(SCD) [Internet]. South San Francisco CA: GBT; 2021 Jul 22 cited Oct
15, 2021]. Available from 2021 Oct 15: https://ir.gbt.com/news-
releases/news-release-details/gbt-provides-regulatory-and-pipeline
-updates-sickle-cell-disease
15. Morrison M, Palermo G, Schmitt C. Lack of ethnic differences in the
pharmacokinetics and pharmacodynamics of inclacumab in healthy
Japanese and Caucasian subjects. Eur J Clin Pharmacol. 2015;71
(11):1365–1374.
16. Tardif JC, Tanguay JF, Wright SR, et al. Effects of the P-selectin antagonist
inclacumab on myocardial damage after percutaneous coronary inter-
vention for non-ST-segment elevation myocardial infarction: results of
the SELECT-ACS trial. J Am Coll Cardiol. 2013;61(20):2048–2055.
17. A study of a single dose of inclacumab to reduce re-admission in
participants with sickle cell disease and recurrent vaso-occlusive crises:
clinicalTrials.gov identifier: NCT04927247; [updated 2021 November 2;
cited 2021 December 3]. Available from 2021 Jun 15: https://clinical
trials.gov/ct2/show/NCT04927247
18. A study to assess the safety and efficacy of inclacumab in partici-
pants with sickle cell disease experiencing vaso-occlusive crises:
clinicalTrials.gov identifier: NCT04935879; [updated
2021 November 5; cited 2021 December 3]. Available from 2021
Jun 23: https://clinicaltrials.gov/ct2/show/NCT04935879
19. Ambrogelly A, Gozo S, Katiyar A, et al., Analytical comparability
study of recombinant monoclonal antibody therapeutics. MAbs.
2018. 10(4): 513–538.
•• Ambrogelly et al provides a review of the current understand-
ing of various modifications of recombinant monoclonal anti-
bodies and outlines the critical steps in designing and
executing successful comparability studies to support process
changes at different stages of a product’s lifecycle.
20. GBT expands sickle cell disease pipeline with worldwide licen-
sing agreement for inclacumab for the treatment of
vaso-occlusive crisis [Internet]. South San Francisco CA: GBT;
2018 Aug 23 cited 2021 Dec 2]. Available from 2021 Dec 2:
https://ir.gbt.com/news-releases/news-release-details/gbt-
expands-sickle-cell-disease-pipeline-worldwide-licensing
21. ICH Q5E: comparability of biotechnological/biological products
subject to changes in their manufacturing process 2005 cited
2021 Dec 2]. Available from 2021 Dec 2: https://www.ema.europa.
eu/en/documents/scientific-guideline/ich-q-5-e-comparability-
biotechnological/biological-products-step-5_en.pdf
22. Guideline on development, Production, characterisation and
specification for monoclonal antibodies and related products;
EMA/CHMP/BWP/532517/2008 2016 cited 2021 Dec 2].
Available from 2021 Dec 2: https://www.ema.europa.eu/en/
documents/scientific-guideline/guideline-development-
production-characterisation-specification-monoclonal-
antibodies-related_en.pdf
23. Guideline on comparability of biotechnology-derived medicinal
products after a change in the manufacturing process. Non-
clinical and clinical issues; EMEA/CHMP/BMWP/101695/2006
2007 cited 2021 Dec 2]. Available from 2021 Dec 2: https://
www.ema.europa.eu/documents/scientific-guideline/guideline-
comparability-biotechnology-derived-medicinal-products-after-
change-manufacturing-process_en.pdf
24. Center for Drug Evaluation and Research. Guidance for Industry.
Statistical Approaches to Establishing Bioequivalence 2001 cited
2022 Oct 26]. Available from 2022 Oct 26: https://www.fda.gov/
media/70958/download
25. ICH Q6B: specifications: test procedures and acceptance criteria for
biotechnological/biological products: ICH; 1999 cited 2021 Dec 3].
Available from 2021 Dec 3: https://database.ich.org/sites/default/
files/Q6B%20Guideline.pdf
26. Bas M, Terrier A, Jacque E, et al. Fc sialylation prolongs serum
half-life of therapeutic antibodies. J Immunol. 2019;202
(5):1582–1594.
27. Ghaderi D, Taylor RE, Padler-Karavani V, et al. Implications of
the presence of N-glycolylneuraminic acid in recombinant ther-
apeutic glycoproteins. Nat Biotechnol. 2010;28(8):863–867.
28. Shields RL, Lai J, Keck R, et al. Lack of fucose on human IgG1
N-linked oligosaccharide improves binding to human Fcgamma
RIII and antibody-dependent cellular toxicity. J Biol Chem.
2002;277(30):26733–26740.
1428 R. MIHAILA ET AL.
