immune pharmacodynamic responses of the novel cancer ... · imprime pgg (imprime) is an i.v....

16
The Journal of Immunology Immune Pharmacodynamic Responses of the Novel Cancer Immunotherapeutic Imprime PGG in Healthy Volunteers Nandita Bose,* Nadine R. Ottoson,* ,1 Xiaohong Qiu,* ,1 Ben Harrison,* Jamie R. Lowe,* Mark T. Uhlik,* Blaine T. Rathmann,* Takashi O. Kangas,* Lindsay R. Jordan,* Kathleen E. Ertelt,* Adria Bykowski Jonas,* Richard M. Walsh,* Anissa S. H. Chan,* Ross B. Fulton,* Steven M. Leonardo,* Kathryn A. Fraser,* Keith B. Gorden,* Mark A. Matson, Jeremy R. Graff,* and Richard D. Huhn* Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors. Imprime-mediated innate immune activation requires immune complex formation with naturally occurring IgG anti-b glucan Abs (ABA). We administered Imprime to healthy human volunteers to assess the necessity of ABA for Imprime-mediated immunopharmacodynamic (IPD) changes. Imprime (4 mg/kg) was administered i.v. in single and multiple infusions. Subsets of subjects were premedicated with antihistamine and corticosteroid. Peripheral blood was measured before, during and after Imprime administration for IPD changes (e.g., ABA, circulating immune complexes, complement activation, complete blood counts, cytokine/chemokine, and gene expression changes). IPD changes were analyzed based on pretreatment serum ABA levels: low-ABA (<20 mg/ml), mid-ABA (20–50 mg/ml), and high-ABA (50 mg/ml). At the end of infusion, free serum ABA levels decreased, circulating immune complex levels increased, and complement activation was observed. At 1–4 h after end of infusion, increased expression of cytokines/chemokines, a 1.5-4-fold increase in neutrophil and monocyte counts and a broad activation of innate immune genes were observed. Low-ABA subjects typically showed minimal IPD changes except when ABA levels rose above 20 mg/ml after repeated Imprime dosing. Mild-to-moderate infusion-related reactions occurred in subjects with ABA 20 mg/ml. Premedications alleviated some of the infusion-related reactions, but also inhibited cytokine responses. In conclusion, ABA levels, being critical for Imprime-mediated immune activation may provide a plausible, mechanism-based biomarker to identify patients most likely to respond to Imprime-based anticancer immunotherapy. The Journal of Immunology , 2019, 202: 000–000. I mprime is a yeast-derived soluble b-1,3/1,6 glucan (BTH1677) that is currently in clinical development for cancer as an i.v. administered immunotherapy in combination with immune checkpoint inhibitor Abs for the treatment of cancer. In previous preclinical and randomized phase II clinical studies, Imprime has demonstrated the potential to enhance the efficacy of tumor targeting and antiangiogenic Abs. In several xenograft models, Imprime enhanced the antitumor efficacy of the anti-EGFR tumor- targeting Ab cetuximab as well as the anti-VEGF Ab bevacizumab (1–3). In a randomized phase II trial, compared with the cohort of advanced nonsquamous non–small cell lung cancer (NSCLC) patients treated with bevacizumab, carboplatin, and paclitaxel, first-line treatment with Imprime plus bevacizumab, carboplatin, and paclitaxel yielded an increased overall response rate and improved median overall survival (4). In high-risk chronic lym- phocytic leukemia patients (including those with del 17p, del 11q risk factors), the addition of Imprime to rituximab and alemtu- zumab yielded a complete response rate of 65%. In a similar trial run at the same institution during the same time period, patients receiving rituximab and alemtuzumab achieved a 36% complete response (5, 6). Mechanistic studies in preclinical tumor models have demon- strated the critical role of complement protein C3, complement receptor 3 (CR3), and Gr1-positive myeloid immune cells in the antitumor activity of Imprime (7, 8). More recently, in vivo mouse studies have also highlighted the need for Dectin-1 re- ceptor in Imprime-mediated innate immune functionality (9). Ex vivo human whole blood studies also showed that Imprime *Biothera Pharmaceuticals, Inc., Eagan, MN 55121; and Prism Research, Saint Paul, MN 55114 1 N.R.O. and X.Q. contributed equally to this work. ORCIDs: 0000-0002-9025-6805 (N.B.); 0000-0002-0276-1474 (T.O.K.); 0000-0002- 4656-8078 (S.M.L.); 0000-0003-3848-7937 (R.D.H.). Received for publication November 28, 2018. Accepted for publication March 14, 2019. This work was supported by Biothera Pharmaceuticals, Inc. Conception and design: R.D.H., J.R.L., N.B., N.R.O., B.H., M.T.U., and J.R.G. Develop- ment of methodology: R.D.H., J.R.L., N.B., N.R.O., and B.H. Acquisition of data (pro- vided animals, acquired and managed patients, provided facilities, etc.): M.A.M., N.R.O., B.H., L.R.J., T.O.K., K.E.E., S.M.L., R.B.F., K.A.F., A.B.J., and K.B.G. Analysis and interpretation of data (e.g. statistical analysis, biostatistics, computational analysis): M.A.M., R.D.H., N.B., N.R.O., A.B.J., B.H., B.T.R., K.E.E., X.Q., and R.M.W. Writing, review, and/or revision of the manuscript: N.B., R.D.H., N.R.O., X.Q., M.T.U., and J.R.G. Administrative, technical, or material support (i.e., reporting or organizing data, con- structing databases): B.H., X.Q., A.S.H.C., and N.R.O. Study supervision: M.A.M., N.B., R.D.H., and J.R.L. All authors read and approved the final manuscript. Address correspondence and reprint requests to Dr. Nandita Bose, Biothera Phar- maceuticals, Inc., 3388 Mike Collins Drive, Eagan, MN 55121. E-mail address: [email protected] The online version of this article contains supplemental material. Abbreviations used in this article: ABA, anti–b-glucan Ab; AE, adverse event; CBC, complete blood count; CH50, total complement activity test; CIC, circulating im- mune complex; CR2, complement receptor 2; CTCAE, Common Terminology Cri- teria for Adverse Events; EOI, end of infusion; IPD, immunopharmacodynamic; NSCLC, non–small cell lung cancer; PAMP, pathogen-associated molecular pattern; SC5b-9, serum complement membrane attack complex. Copyright Ó 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1801533 Published April 15, 2019, doi:10.4049/jimmunol.1801533 by Nandita Bose on April 15, 2019 http://www.jimmunol.org/ Downloaded from

Upload: others

Post on 07-Jul-2020

5 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

The Journal of Immunology

Immune Pharmacodynamic Responses of the Novel CancerImmunotherapeutic Imprime PGG in Healthy Volunteers

Nandita Bose,* Nadine R. Ottoson,*,1 Xiaohong Qiu,*,1 Ben Harrison,* Jamie R. Lowe,*

Mark T. Uhlik,* Blaine T. Rathmann,* Takashi O. Kangas,* Lindsay R. Jordan,*

Kathleen E. Ertelt,* Adria Bykowski Jonas,* Richard M. Walsh,* Anissa S. H. Chan,*

Ross B. Fulton,* Steven M. Leonardo,* Kathryn A. Fraser,* Keith B. Gorden,*

Mark A. Matson,† Jeremy R. Graff,* and Richard D. Huhn*

Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors.

Imprime-mediated innate immune activation requires immune complex formation with naturally occurring IgG anti-b glucan

Abs (ABA). We administered Imprime to healthy human volunteers to assess the necessity of ABA for Imprime-mediated

immunopharmacodynamic (IPD) changes. Imprime (4 mg/kg) was administered i.v. in single and multiple infusions. Subsets of

subjects were premedicated with antihistamine and corticosteroid. Peripheral blood was measured before, during and after

Imprime administration for IPD changes (e.g., ABA, circulating immune complexes, complement activation, complete blood

counts, cytokine/chemokine, and gene expression changes). IPD changes were analyzed based on pretreatment serum ABA levels:

low-ABA (<20 mg/ml), mid-ABA (‡20–50 mg/ml), and high-ABA (‡50 mg/ml). At the end of infusion, free serum ABA levels

decreased, circulating immune complex levels increased, and complement activation was observed. At ∼1–4 h after end of

infusion, increased expression of cytokines/chemokines, a 1.5-4-fold increase in neutrophil and monocyte counts and a broad

activation of innate immune genes were observed. Low-ABA subjects typically showed minimal IPD changes except when ABA

levels rose above 20 mg/ml after repeated Imprime dosing. Mild-to-moderate infusion-related reactions occurred in subjects with

ABA ‡20 mg/ml. Premedications alleviated some of the infusion-related reactions, but also inhibited cytokine responses. In

conclusion, ABA levels, being critical for Imprime-mediated immune activation may provide a plausible, mechanism-based

biomarker to identify patients most likely to respond to Imprime-based anticancer immunotherapy. The Journal of Immunology,

2019, 202: 000–000.

Imprime is a yeast-derived soluble b-1,3/1,6 glucan (BTH1677)that is currently in clinical development for cancer as an i.v.administered immunotherapy in combination with immune

checkpoint inhibitor Abs for the treatment of cancer. In previouspreclinical and randomized phase II clinical studies, Imprime hasdemonstrated the potential to enhance the efficacy of tumortargeting and antiangiogenic Abs. In several xenograft models,Imprime enhanced the antitumor efficacy of the anti-EGFR tumor-targeting Ab cetuximab as well as the anti-VEGFAb bevacizumab(1–3). In a randomized phase II trial, compared with the cohortof advanced nonsquamous non–small cell lung cancer (NSCLC)patients treated with bevacizumab, carboplatin, and paclitaxel,first-line treatment with Imprime plus bevacizumab, carboplatin,and paclitaxel yielded an increased overall response rate and

improved median overall survival (4). In high-risk chronic lym-phocytic leukemia patients (including those with del 17p, del 11qrisk factors), the addition of Imprime to rituximab and alemtu-zumab yielded a complete response rate of 65%. In a similar trialrun at the same institution during the same time period, patientsreceiving rituximab and alemtuzumab achieved a 36% completeresponse (5, 6).Mechanistic studies in preclinical tumor models have demon-

strated the critical role of complement protein C3, complementreceptor 3 (CR3), and Gr1-positive myeloid immune cells in theantitumor activity of Imprime (7, 8). More recently, in vivomouse studies have also highlighted the need for Dectin-1 re-ceptor in Imprime-mediated innate immune functionality (9).Ex vivo human whole blood studies also showed that Imprime

*Biothera Pharmaceuticals, Inc., Eagan, MN 55121; and †Prism Research, Saint Paul,MN 55114

1N.R.O. and X.Q. contributed equally to this work.

ORCIDs: 0000-0002-9025-6805 (N.B.); 0000-0002-0276-1474 (T.O.K.); 0000-0002-4656-8078 (S.M.L.); 0000-0003-3848-7937 (R.D.H.).

Received for publication November 28, 2018. Accepted for publication March 14,2019.

This work was supported by Biothera Pharmaceuticals, Inc.

Conception and design: R.D.H., J.R.L., N.B., N.R.O., B.H., M.T.U., and J.R.G. Develop-ment of methodology: R.D.H., J.R.L., N.B., N.R.O., and B.H. Acquisition of data (pro-vided animals, acquired and managed patients, provided facilities, etc.): M.A.M., N.R.O.,B.H., L.R.J., T.O.K., K.E.E., S.M.L., R.B.F., K.A.F., A.B.J., and K.B.G. Analysis andinterpretation of data (e.g. statistical analysis, biostatistics, computational analysis):M.A.M., R.D.H., N.B., N.R.O., A.B.J., B.H., B.T.R., K.E.E., X.Q., and R.M.W. Writing,review, and/or revision of the manuscript: N.B., R.D.H., N.R.O., X.Q., M.T.U., and J.R.G.

Administrative, technical, or material support (i.e., reporting or organizing data, con-structing databases): B.H., X.Q., A.S.H.C., and N.R.O. Study supervision: M.A.M.,N.B., R.D.H., and J.R.L. All authors read and approved the final manuscript.

Address correspondence and reprint requests to Dr. Nandita Bose, Biothera Phar-maceuticals, Inc., 3388 Mike Collins Drive, Eagan, MN 55121. E-mail address:[email protected]

The online version of this article contains supplemental material.

Abbreviations used in this article: ABA, anti–b-glucan Ab; AE, adverse event; CBC,complete blood count; CH50, total complement activity test; CIC, circulating im-mune complex; CR2, complement receptor 2; CTCAE, Common Terminology Cri-teria for Adverse Events; EOI, end of infusion; IPD, immunopharmacodynamic;NSCLC, non–small cell lung cancer; PAMP, pathogen-associated molecular pattern;SC5b-9, serum complement membrane attack complex.

Copyright� 2019 by TheAmerican Association of Immunologists, Inc. 0022-1767/19/$37.50

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1801533

Published April 15, 2019, doi:10.4049/jimmunol.1801533 by N

andita Bose on A

pril 15, 2019http://w

ww

.jimm

unol.org/D

ownloaded from

Page 2: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

triggers a myriad of innate immune responses, including com-plement activation, neutrophil and monocyte binding and pheno-typic activation, cytokine/chemokine production, and enhancedmyeloid cell-mediated tumor cell killing, in concert with antitu-mor Abs (10, 11). Interestingly, Imprime binding and functionalactivation of these innate effector cells was critically dependent onImprime first forming an immune complex with endogenous IgGanti–b-glucan Ab (ABA). The Imprime–ABA immune complexactivates the classical complement pathway, becomes opsonizedby complement protein fragments, and can then activate innateeffector functionality through CR3 and FcgRIIa (10, 11). Donorblood from healthy subjects with insufficient ABA levels neitherbound nor functionally responded to Imprime unless rescued bythe addition of purified ABA or commercial IVIG, which is richin ABA (11). Furthermore, retrospective analyses of historicImprime-based cancer clinical studies have revealed that patientswith higher pretreatment ABA IgG levels derived more clinicalresponse benefit from Imprime-based therapy than patients withlow ABA levels (12).In the present phase I healthy volunteer study, we sought in vivo

evidence for the relationship between ABA levels and Imprime-mediated innate immune activation. We monitored several innateimmune activation effects in 30 subjects with different pretreat-ment serum ABA levels dosed with 4 mg/kg, the standard dose ofImprime in clinical trials. Consistent with our ex vivo findings,Imprime dosing resulted in several innate immune activatingpharmacodynamic changes that were measurable at both serum/plasma and the cellular levels. These changes included an acutedrop in the free ABA levels concomitant with increased levels ofcirculating immune complex (CIC), complement activation, pro-duction of cytokines and chemokines, expansion of key anticancereffector cells, neutrophils and monocytes, and changes in theexpression of a broad set of early innate immune response genes.Notably, these pharmacodynamic changes were minimal in thelow-ABA group (,20 mg/ml). Furthermore, the adverse event(AE) profile of Imprime essentially mirrored the immuno-pharmacodynamic (IPD) profile with significantly lower AE in thelow-ABA group. Finally, a premedication regimen that includedcorticosteroids abrogated the cytokine response to Imprime.Collectively, these data confirm the critical role of endogenous

ABA levels in Imprime-mediated innate immune activation in theclinical setting and also support the use of ABA levels for patientselection in ongoing cancer clinical trials.

Materials and MethodsTrial design and drug administration

The clinical protocol and informed consent forms were reviewed and ap-proved by Salus institutional review board (Austin, TX). The study wasconducted in three parts as shown in the schema (Fig. 1). Cohort 1(12 subjects) received a single dose of Imprime 4 mg/kg i.v. over 2 or 3 h(depending on volume of administration) 1 wk following an infusion of asimilarly calculated volume of 0.9% normal saline. Cohort 2 (12 subjects)received an infusion of 0.9% normal saline followed by three weekly i.v.infusions of 4 mg/kg Imprime. Six of the 12 subjects in each cohort 1 and 2were assigned to receive premedications with low-dose corticosteroids(8 mg of dexamethasone) and a low-dose H1 antagonist (50 mg diphen-hydramine i.v.) 1 h before dosing with Imprime. The remaining sixsubjects in each cohort 1 and 2 did not receive premedication. Thenonpremedicated subjects who could not complete a prior dose infusion(e.g., because of AE) could be redosed on schedule with premedication onthe subsequent dose. Cohort 3 (six subjects) received i.v. infusions of4 mg/kg on the first, second, and fifth weeks without premedication.Subjects were observed for AE by clinic staff, including a physicianduring infusions and for 8 h following infusions. AE were classified andgraded by the clinical physician according to the National Institutes ofHealth/National Cancer Institute Common Terminology Criteria for Ad-verse Events (CTCAE) version 4.03 (13). The demographics and ABA

values for all the subjects who did not receive and those who receivedpremedications are listed in Table I.

All samples were collected in accordance with Salus independent reviewboard–approved protocol and the New England independent review board–approved protocol. All patients provided written consent prior to bloodcollection.

All data generated or analyzed in this study that are relevant to the resultspresented in this article are included in this article and its supplementaryinformation files. Other data that were not relevant to the results presentedhere are available from the corresponding author upon reasonable request.

Laboratory evaluations

Subjects were healthy volunteers as assessed by medical and social history,physical examination, and screening laboratory testing including hepatitis Band C, HIV, serum chemistry, coagulation tests, complete blood count(CBC) with differential, serum pregnancy (b-HCG) test, urinalysis, andelectrocardiogram. Throughout the study, subjects were monitored byphysical examination, vital signs, serum chemistry, coagulation, and AEs.

IPD evaluations

Human serumwas prepared using Vacutainer SST tubes (Becton Dickinson,Franklin Lakes, NJ) according to the manufacturer’s instructions, andthe cleared serum samples were used immediately or stored at –80˚C.Serum samples were tested for ABA and CIC formation at the followingtimepoints: predose, approximate end of infusion (EOI), 24, and 168 hpost-EOI. Complement activation was measured at predose, ∼EOI, and24 h post-EOI in plasma specimens. Serum chemokine and cytokineproduction were measured at predose, ∼EOI,∼1–4 h post-EOI, and 24 hpost-EOI. CBC with differential were assessed predose, 15 min to 1 h afterstart of infusion, ∼EOI, and ∼4, 24, and 168 h post-EOI. Gene expressionwas quantified from whole blood samples collected predose, ∼1, and 24 hpost-EOI.

ELISA for ABA measurement

To measure the concentrations of serum IgG and IgM ABA, solid-phasedirect ELISAs were developed. Diluted serum was incubated on a micro-titer plate precoated with Imprime. The plates were then incubated with eitheranti-human g-chain or anti-human m chain Abs conjugated to peroxi-dase (Jackson ImmunoResearch, West Grove, PA). A chromogenicsubstrate (SeraCare, Milford, MA) was subsequently added, and theresulting color change was measured spectrophotometrically at ab-sorbance of 450 nm. A standard curve based off of purified human ABA

FIGURE 1. Healthy volunteer clinical study schema. Twelve individu-

als in cohort 1 (six non-premedicated and six premedicated with diphen-

hydramine 50 mg i.v. and dexamethasone 25 mg i.v. 1 h before dosing)

received 0.9% normal saline in week 1 followed by Imprime in week 2.

Twelve subjects in cohort 2 (six non-premedicated and six premedicated as

in cohort 1) received 0.9% normal saline (NS) in week 1 followed by three

weekly doses of Imprime. Six subjects in cohort 3 subjects received no

premedication and were dosed with Imprime in weeks 1, 2, and 5.

2 ANTI–b-GLUCAN Ab–DEPENDENT PHARMACODYNAMICS OF IMPRIME PGG

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 3: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

was run on each assay to determine microgram per milliliter concentra-tions of samples.

CIC measurement

CIC levels were measured by using MicroVue Complement CIC-Raji CellReplacement Kit (Quidel Corporation, San Diego, CA) with SpectraMax-M5 plate reader (Molecular Devices, San Jose, CA).

Complement measurement

The classical complement activity in plasma was measured by usingthe MicroVue Complement CH50 Eq EIA Kit (Quidel Corporation), andthe complement levels of serum complement membrane attack complex(SC5b-9) were measured by using MicroVue SC5b-9 Plus EIA Kit (QuidelCorporation).

Cytokine and chemokine measurement

All or a subset of the cytokines and chemokines, including eotaxin,GM-CSF, IFN-g, IFN-a1, IL-1b, IL-1RA, IL-10, IL-12, IL-13, IL-15,IL-17, IL-2, IL-2R, IL-4, IL-5, IL-6, IL-7, IL-8, IP-10 (CXCL10),MCP-1 (CCL2), MIG (CXCL9), MIP-1a (CCL3), MIP-1b (CCL4),RANTES (CCL5), and TNF-a, were tested using multiplex or sin-gleplex Luminex kits (EMD Millipore, Billerica, MA and Thermo-Fisher Scientific, Pittsburg, PA) at various time points before and

after Imprime infusion. Results were obtained on Luminex xMAP200 (ThermoFisher Scientific) using xPonenent software and ana-lyzed on Milliplex Analyst software. For ex vivo experiments per-formed to test the effect of dexamethasone on Imprime-inducedcytokine production, whole blood from individuals with high ABA(60–210 mg/ml) was treated ex vivo with vehicle alone or Imprime at25 mg/ml in the presence of dexamethasone (0.2 and 2 mM) over24 h. Plasma was collected by centrifugation and assayed byLuminex.

Complete blood counts

Automated CBC with differential counts were performed by a ClinicalLaboratory Improvement Amendments–certified contract laboratory at thetime points mentioned above (MedTox Laboratories, St. Paul, MN).

Transcriptional regulation

Human blood samples from various time points before and afterImprime infusion were collected using PAXgene blood RNA tube (BDBiosciences, Franklin Lakes, NJ) according to the manufacturer’s in-structions and stored at –80˚C. RNA was isolated with RNeasy MiniKit (Qiagen, Hilden, Germany). A custom-made 80-plex QuantiGeneassay kit (ThermoFisher Scientific) using a bead-based multiplex assaywas used to measure the expression of 80 genes of interest, including72 target genes and 8 reference genes with Luminex xMAP 200

Table I. Demographics and baseline ABA values of healthy volunteers (low-ABA, mid-ABA, and high-ABA) administered Imprime with and withoutpremedications

Red indicates low-ABA, black indicates mid-ABA, and blue indicates high-ABA.*Discontinued from the study and therefore not included in the analyses of IPD responses.**Received an incomplete first dose of Imprime without premedications but received the subsequent two doses along with premedications.†Received premedications with the third dose of Imprime.BMI, body mass index; F, female; M, male.

The Journal of Immunology 3

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 4: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

according to the manufacturer’s instructions. Gene expression valueswere normalized relative to geometric mean of six housekeeping genes.Fold changes relative to preinfusion control were log2 transformed andpresented as a heatmap.

ResultsTo isolate the IPD effects of Imprime alone from those of thepremedications, the pharmacodynamic responses of Imprimewere assessed in all the subjects across the three cohorts whoreceived a full 4 mg/kg dose Imprime without any premed-ications (Fig. 1). These subjects were divided into three groupsbased on their pre-Imprime IgG ABA values (Table I): low-ABA (,20 mg/ml), mid-ABA ($20–50 mg/ml), and high-ABA ($50 mg/ml). The IPD responses to a single dose andrepeated doses measured in this study were changes in the ABAlevels, formation of CIC, complement activation, cytokineproduction, CBC with differentials, and whole blood gene ex-pression levels. As controls, some of these IPD changes werealso measured in a subset of subjects in response to salineadministration.

Imprime administration results in decreased levels of freeABA with a concurrent increase in IgG-containing CIC inthe serum

We hypothesized that Imprime administration should result inImprime–ABA immune complex formation and, as a result,should cause a measurable drop in the serum levels of free IgGABA as well as an increase in the levels of CIC. Subsequent re-sults showed the free ABA levels were indeed reduced at the EOIin each of the three groups of subjects with the median drop of22% (range 0.6- to 0.9-fold) in the low-ABA group, 33% mediandrop (range 0.5- to 0.9-fold) in the mid-ABA group, and 50% drop(range 0.4- to 0.6-fold) in the high-ABA group. The levelsremained reduced at 24 h before returning to baseline between 24and 168 h postdose in most subjects. Interestingly, several sub-jects in each of the three groups showed increases in ABA levelsranging from ∼1.5- to 6.1-fold increase over baseline by 168 h(Fig. 2A). Accompanying the drop in the free ABA levels, we sawan increase in the CIC levels with median 1.2-fold (range 0.9–1.8)in the low-ABA group, median 2.0-fold (range 1.3–3.7) in themid-ABA group, and median 3.9-fold (range 2.3–4.5) in the high-ABA group (Fig. 2B). The peak levels were achieved betweenEOI to 24 h post-EOI and returned to baseline by 168 h. As theserum CIC levels do not account for the immune complexes thatare cell-bound, they do not reflect the total CIC formed afterImprime administration. Additionally, the ELISA used to measure

ABA does not capture and detect ABA sequestered in the immunecomplex. Although Imprime-specific CIC were not measured, thelack of drop in the free ABA levels or increase in the CIC levels insubjects who only received saline (data not shown) does provideindirect in vivo evidence of the Imprime–ABA immune complexformation.

Imprime administration results in complement activation

As the Imprime–ABA immune complex has been shown to acti-vate the classical complement pathway and produce complementactivation products, C4a, C5a, and SC5b-9, we surmised thatImprime, upon formation of the Imprime–ABA complex, wouldfix the complement protein C3. This reaction should consume C3and then result in reduction of the total complement activity asmeasured by the total complement activity test (CH50) test. Wetherefore measured the SC5b-9 levels and CH50 activity from theplasma collected at preinfusion and ∼EOI. A median of 1.8-foldincrease (range 1.2–3.1) in the SC5b-9 levels and no change inthe CH50 values were observed in the low-ABA group. A median4.8-fold increase (range 1.9–28.0) in SC5b-9 levels and a mediandrop of 32% (range 1.1- to 0.7-fold) in CH50 activity were ob-served in the mid-ABA group. The high-ABA group showed thehighest increase in the SC5b-9 levels and decrease in CH50 activitywith a median of 17.8-fold increase (range 8.8–103.9) and 53%median drop (range 0.8- to 0.1-fold), respectively (Fig. 3). Comple-ment activation changes were not evident with saline administra-tion alone (data not shown), confirming the specificity of Imprime’sability to activate complement in an ABA-dependent manner.

Imprime induces a robust increase in systemic cytokinesand chemokines

Ex vivo human whole blood studies have shown that Imprimeconsistently induces chemokines, including IL-8 and MCP-1, inblood of subjects with sufficiently high ABA levels (11). In thistrial, multiplex evaluation of serum cytokines/chemokines showedthat Imprime administration in the high-ABA group elicited.2-fold increase above predose in IL-8 (four of four subjects),MCP-1 (four of four subjects), MIP-1a (three of four subjects),MIP-1b (four of four subjects), IL-6 (four of four subjects), IP-10(three of four subjects), TNF-a (three of four subjects), and IL-10(two of four subjects). The cytokine/chemokine levels peaked be-tween EOI and 4 h after EOI before returning to baseline by 24 h. Inthe mid-ABA group, the cytokine/chemokine production was moresporadic, with two of seven subjects producing some of the afore-mentioned cytokines/chemokines. Imprime did not stimulate anysignificant production of the measured cytokines/chemokines in the

FIGURE 2. Imprime administration

resulted in drop of free IgG ABA con-

centrations and increase in CIC levels.

Serological changes in IgG ABA and

CIC levels with a single dose of Imprime

were tested at predose, ∼EOI, ∼24, and168 h post-EOI using an ELISA method.

The change in absolute concentrations

and folds over time and maximum fold

change at ∼EOI are shown for (A) IgG

ABA and (B) CIC in low-, mid, and

high-ABA groups. Each line or point

represents an individual subject in each

of the ABA groups.

4 ANTI–b-GLUCAN Ab–DEPENDENT PHARMACODYNAMICS OF IMPRIME PGG

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 5: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

low-ABA group at any of the time points (Fig. 4). Saline admin-istration also did not induce any cytokine/chemokine responses inany of the ABA groups (data not shown).

Imprime induces expansion of neutrophils and monocytes inthe blood

We next tested whether Imprime administration led to any changesin the absolute numbers of neutrophils, monocytes, and lymphocytesin the blood by performing the CBC with differentials. In the high-ABA group, Imprime mediated a rapid increase in the absolutenumbers of neutrophils and monocytes at ∼EOI through 4 h post-EOI before returning to near baseline by 24 h. A median 3.1-foldincrease in neutrophils (range 2.4–3.4) and 1.6-fold increase in themonocyte count (range 1.3–2.1) were observed (Fig. 5). When welooked at cell count changes at earlier time points (i.e., duringImprime infusion) in a subset of these subjects, we found thatImprime first induced rapid neutropenia and monocytopenia asearly as 15 min into the infusion before the expansion at latertime points (Supplemental Fig. 1). In contrast, lymphopeniawithout any further expansion was observed in these subjects,with a median of 65% drop in the lymphocyte count (range 0.3-to 0.4-fold). The leukocyte counts in most of these subjects

returned to near baseline in ∼24 to 48 h. In the mid-ABA group,the increases in neutrophil and monocyte counts were modest,with median increases of 1.6-fold (range 1.3–2.9) and 1.5-fold(range 1.3–2.3), respectively. Additionally, a 24% median de-crease was seen in the lymphocyte counts. There were no changesin the leukocyte counts in the low-ABA group. These alterationsin leukocyte counts were not observed with saline administrationalone (data not shown). These results again highlighted the abilityof Imprime to induce changes in the myeloid cell count in thosewith ABA $20 mg/ml.

Imprime induces a transcriptional signature of innate immuneactivation in blood

To determine a broad molecular signature of Imprime, we firstevaluated the expression of selected gene transcripts in wholeblood. This investigation was done as part of cohort 3 only and, assuch, had one subject in the low-ABA group, two from the mid-ABA group, and two from the high-ABA group. The genes forwhich the mRNA transcription was differentially changed (i.e., upor down) from baseline by at least 2-fold are presented as a log2-transformed gene expression heat map (Fig. 6). In the high-ABAgroup, both subjects substantially upregulated (∼2- to 97-fold)several innate immune activation markers: CD83, ICAM1, CD274(PD-L1), ITGAM (CD11b), FCGR1B, FCGR2A, FCGR3A,FCGR3B, and C5AR1; chemokine/chemokine receptor family ofgenes: CXCL1, CXCL2, CXCL8, CCL2, CCL3, CCL4, CXCR1,CXCR4, and CCR1; and cytokine genes: IL1B, IL1RN, IL10, andTNFA. CCL5 and CCR5 were the only measured chemokine/receptor pair to show downregulation. Among the type I IFN–related genes tested, a 2-fold increase in the IFN-stimulated geneMX2 and a decrease in the OAS2 gene were seen in both sub-jects. Interestingly, among the B cell markers evaluated, SDCBP(CD138), which is primarily expressed on Ab secreting plasmacells, was highly upregulated in both the high-ABA donors. Themid-ABA subjects showed a modest (2- to 4-fold) increase in asubset of the genes upregulated in the high-ABA subjects. In boththe high- and mid-ABA groups, maximum upregulation of thegenes occurred at about ∼1 h post-EOI and then returned tobaseline by ∼24 h. Consistent with the earlier IPD findings,Imprime did not significantly induce any gene expression changesin the low-ABA subject.

IPD effects of repeated Imprime doses

We next evaluated whether repeated dosing of Imprime altered anyof the IPD effects. Ten subjects received multiple doses of Imprime.As shown previously in Fig. 2, some subjects in each of the threeABA groups showed increased IgG ABA levels after one dose ofImprime. Examination of changes in the IgM ABA levels found

FIGURE 4. Imprime induces cytokine production. Serum cytokine levels produced by one dose of Imprime at ∼1–4 h post-EOI were measured by

Luminex multiplex assay. The maximum fold change over predose values are plotted for each of the subjects in (A) low-ABA, (B) mid-ABA, and (C) high-

ABA groups. Each line represents an individual subject in each of the ABA groups.

FIGURE 3. Imprime administration results in complement activation.

Complement activation in response to a single dose of Imprime was

measured as changes in plasma levels of SC5b-9 and CH50 activity at

predose and ∼EOI using commercial enzyme immunoassay kits. The

changes in (A) SC5b-9 levels and (B) CH50 activity are shown as absolute

concentrations and maximum fold change. Each line or point represents an

individual subject in each of the ABA groups.

The Journal of Immunology 5

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 6: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

that subjects with increased IgG levels did not show concomitantincrease in the IgM levels (Fig. 7A). This increase in IgG ABAwas generally associated with an enhanced immunological re-sponse to Imprime and was most striking in the low-ABA group.Some of these subjects were not responsive to the first dose ofImprime but showed robust responsiveness by the third dose ofImprime, coinciding with increased ABA levels. Representativedata presented for one low-ABA subject show that as the ABAlevels increased from 15.0 mg/ml (predose 1) to 26.2 mg/ml(predose 3), IPD responses were observed after the third dose ofImprime. A 2.7-fold increase in CIC formation, a 4.9-fold increasein SC5b-9 production, .2-fold increase in cytokine/chemokinesIL-6, IL-8, MCP-1, MIP-1a, MIP-1b, MIG, IP-10, and TNF-a,70% drop followed by 3.2-fold increase in the neutrophil counts(data not shown for monocytes), and .2-fold change in the ex-pression of several innate immune genes were observed (Fig. 7B).Interestingly, the gene expression of the plasmablast/plasma cellmarker, CD138, also increased ∼3-fold by the third dose ofImprime. Representative data for one subject in each of the mid-and high-ABA groups show that the mid-ABA subject also hadenhanced IPD responses as ABA levels increased 2.6-fold by thethird dose of Imprime, whereas the high-ABA subject with only amodest increase in ABA levels by the third dose had no significantchanges in the responses (Supplemental Fig. 2). The enhancedIPD responses with the increased ABA levels demonstrate thatthese Abs are not neutralizing Abs but, rather, are critical forImprime-mediated innate immune responses.

Effects of premedications on Imprime’s IPD

Medications such as corticosteroids, given concomitantly orprophylactically, have been shown to suppress the activity ofinnate immune cells (14). In our ex vivo experiments, dexa-methasone completely inhibited Imprime-mediated cytokineproduction (Supplemental Fig. 3). Thus, we evaluated the effectof dexamethasone and diphenhydramine, a standard prophylacticpremedication regimen administered occasionally with Imprimeand often with other biologics and chemotherapy, on Imprime-mediated IPD responses. Eleven subjects who received full dosesof Imprime along with premedications were evaluated (Table I).The premedications profoundly inhibited Imprime-mediated

cytokine production. The comparison of mean concentrations ofcytokines produced in the high-ABA subjects who were not pre-medicated (n = 5) versus who were premedicated (n = 7) showsthat consistent with the ex vivo data, the premedications consid-erably inhibited Imprime-mediated cytokine production (Fig. 8A).Furthermore, even with an increase in the ABA levels after

repeated dosing of Imprime, the cytokine levels remained blunted inthese subjects (Fig. 8B). The other pharmacodynamic responses,including Imprime-mediated drop in free ABA levels, increase inCIC formation, and increased complement activation, were pre-served in the mid- and the high-ABA groups (Supplemental Fig. 4).However, consistent with the known leukocytosis effect ofcorticosteroids, increased expansion of neutrophils, monocytes,and lymphocytes were seen in all subjects independent of ABAlevels (15). These results indicate that corticosteroid premed-ication counteracts some of the key innate immune modulatingeffects of Imprime.

Imprime induces infusion-related reactions

As the pharmacodynamic effects of Imprime correlate with theABA levels, we assessed whether there is a relationship betweenImprime-induced AEs and ABA levels. We first evaluated the AEsin subjects who did not receive any premedications. In Table II, theIgG ABA values prior to each Imprime dose along with a desig-nation of whether there was an infusion reaction and its highestCTCAE grade of any symptom component are listed. All symp-toms were mild to moderate in severity (CTCAE grade 1–2), withthe exception of one grade 3 headache. The symptoms typicallyresolved within approximately one-half hour following EOI. Be-cause the ABA values increased in some subjects with repeateddosing of Imprime, we evaluated the AEs based on each of thepredose ABA level. The maximum grade of any componentsymptom associated with any dose of Imprime was plotted againsteach of the predose ABA categories (Fig. 9). Most doses were notassociated with any infusion reaction. There were two grade 1events in the category of doses associated with ,20 mg/ml pre-dose ABA levels. There were four grade 1 events and two grade 2events in the doses associated with 20–50 mg/ml predose ABA

FIGURE 6. Imprime induces transcriptional changes in whole blood.

Whole blood from cohort 3 subjects was collected in PaxGene RNA col-

lection tubes at predose, ∼1, and 24 h post-EOI. RNA was isolated and

assayed by QuantiGene Plex Kit. The genes for which the mRNA tran-

scription was differentially changed from baseline by at least 2-fold are

presented as a log2-transformed gene expression heatmap.

FIGURE 5. Imprime induces change in neutrophils, monocytes, and

lymphocyte counts. CBC values were obtained between ∼15 min after start

of infusion and ∼4 h post-EOI, and maximum fold change over the predose

value was calculated for neutrophils (left panel), monocytes (center panel),

and lymphocytes (right panel). Each point represents an individual subject

in each of the ABA groups.

6 ANTI–b-GLUCAN Ab–DEPENDENT PHARMACODYNAMICS OF IMPRIME PGG

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 7: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

levels. Finally, in the category of doses with $50 mg/ml pre-dose ABA levels, there were five grade 1 events, nine grade 2events, and one grade 3 event. Thus, higher grade AEs appearedto be associated with higher predose ABA levels ($20 mg/ml),albeit high ABA levels did not always result in infusionreactions.We next looked at the severity of the AEs in subjects who re-

ceived premedication (Table III). One high-ABA subject experi-enced a grade 3 AE, but the symptoms rapidly resolved upondiscontinuation of the infusion. No grade 2 AEs were observedin these subjects. Even though some subjects increased theirABA levels by dose 2 or 3, only two subjects developed grade 1AEs with subsequent doses of Imprime. Thus, corticosteroidsmay generally mitigate the severity of Imprime-induced infusionreactions but also substantially blunt cytokine responses to

Imprime and may therefore defeat Imprime-mediated innateimmune responses.

DiscussionIn our previous studies using ex vivo human whole blood, we haveshown that Imprime is pharmacologically active as an immunecomplex formed between Imprime and naturally occurring ABA(10, 11). The concentration of ABA in serum is a critical deter-minant of the formation of this complex, subsequent opsonization,binding to innate immune cells, and innate immune functionalactivation. The role of ABA in mediating the pharmacodynamiceffects, as well as the clinical response to Imprime in the clinicaltrials has not been well established. In this study, we have ex-plored the relationship between the pretreatment ABA levels andthe in vivo IPD effects of Imprime.

FIGURE 7. Imprime-induced IPD

changes with repeated dosing. Serological

changes in ABA levels and IPD changes

with repeated dosing of Imprime were

measured using the methods previously

described. (A) Longitudinal changes in IgG

and IgM ABA levels at predose, ∼EOI, and∼24 h post-EOI for each of the three doses

are presented for the three ABA groups. (B)

A representative example of a low-ABA

subject showing (a) increasing ABA levels

with repeated dosing of Imprime and

thereby enhanced IPD responses as mea-

sured by increased (b) CIC levels, (c)

SC5b-9 levels, (d) cytokine production,

(e) fold change in neutrophil counts, and

(f) gene expression changes.

The Journal of Immunology 7

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 8: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

The systemic effects of Imprime were assessed as serum andcellular immune responses in subjects with varying ABA levels.The formation of in vivo Imprime–IgG ABA immune complex wasevident in the drop in the free IgG ABA levels and the simulta-neous increase in the levels of IgG-containing CIC observed withthe administration of every dose of Imprime (Fig. 2). One of theeffects directly downstream of immune complex formation that isproportional to the existing Ab levels is the activation of theclassical complement pathway. True to this, Imprime adminis-tration resulted in complement fixation, yielding a greater in-crease in the SC5b-9 and reciprocal drop in CH50 levels in themid- and high-ABA group (Fig. 3). Importantly, even thoughhigher levels of SC5b-9 were produced in these subjects, clinicalsymptoms sometimes associated with the deposition of themembrane form of the terminal complement complex (C5b-9)such as hemolytic anemia or pancytopenia were not observed.This is consistent with our earlier reported findings that themembrane bound-C5b-9 was not detected on innate immune

cells, thus indicating that Imprime-induced complement activationis not overwhelming the complement regulatory processes (11).In addition to the immunological changes in the serum, a single

dose of Imprime induced rapid and robust cellular changes in thehigh-ABA group, including induced expression of innate immuneresponse genes in the blood, systemic production of cytokine/chemokines, and expansion of neutrophils and monocytes. Thegene expression profile included a range of innate immune acti-vation targets, including modulation of innate receptors, dendriticcell maturation markers, dendritic cell and monocyte activationmarkers, chemokines and chemokine receptor family, adhesionmolecule, and select proinflammatory and anti-inflammatory genes(Fig. 6). Consistent with the gene expression profile, Imprimestimulated production of several chemokines and only low levelsof select proinflammatory cytokines such as IL-6 and TNF-a.The increase in anti-inflammatory cytokine IL-10 was sporadic(Fig. 4). Considering that Imprime induced key chemokinesknown to be involved in margination/demargination, trans-endothelial migration, and mobilization of myeloid and lymphoidcells (e.g., CXCL1, CXCL2, IL-8, MCP-1, MIP-1a, IP-10), it isnot surprising that there was a rapid, transient reduction in theneutrophil, monocyte, and lymphocyte count followed by sub-stantial expansion of the neutrophil and monocyte counts (Fig. 5)(16, 17). Moreover, the effect of b-glucans on mobilization ofmyeloid cells from the bone marrow is already well documented(18, 19). Thus, Imprime-induced pharmacodynamic serum andcellular changes were consistent with the early innate immuneactivation events elicited by a pathogen or a pathogen-associatedmolecular pattern (PAMP). However, unlike other PAMP agents,including TLR and/or STING, agonists that are precluded fromsystemic clinical administration because of the potential risk ofinducing cytokine storm or immune cell death, Imprime’s uniquecytokine profile allows it to be well tolerated when administeredsystemically (20–22). Although some of these PAMP-based innateimmune modulators when administered in a biomarker-unselectedpatient population have been demonstrated to induce some of thesystemic immune activity, translation to clinical efficacy is yet tobe shown. As Imprime is currently being clinically developed inconcert with the T cell checkpoint inhibitors, anti–PD-1 (pem-brolizumab) in metastatic melanoma and triple-negative breastcancer (NCT02981303) and anti–PD-L1 (atezolizumab) in meta-static colorectal cancer (NCT03555149), some of the innate im-mune changes in both peripheral blood and tumor tissue that arecritical to evoke, expand, and sustain T cell immunity, includingthe production of select chemokines required for T cell migra-tion, together with changes in the expression of costimulatorymarkers on the myeloid APC, are being extensively evaluated(23, 24). As ABA levels are being measured in these trials, it willbe interesting to see the correlation between the ABA levels, theinnate immune pharmacodynamic responses, and ultimately, theclinical responses.Imprime-induced IPD changes were consistently not observed in

the low-ABA subjects. Strikingly though, some of the low-ABAsubjects showed responses to Imprime after their ABA levels in-creased and crossed the threshold of ∼20 mg/ml upon multipleweekly dosing of Imprime (Fig. 7). As Ag–Ab ratio is one of thecritical determinants of immune complex formation and down-stream innate immune functions, it is not surprising that there maybe an optimal threshold of IgG ABA required for Imprime tocomplex with and consequently bind to and activate the innatecells. However, our earlier study had shown that subjects withhigher IgM did not produce some of the cytokines in response toImprime because of lack of FcgRIIA engagement (11). So it ispossible that other factors, including ratio of IgG/IgM ABA or the

FIGURE 8. Premedications inhibit Imprime-induced cytokine produc-

tion. Cytokine levels in subjects who received Imprime with and without

premedications were measured from serum collected ∼1–4 h post-EOI by

Luminex multiplex assay. (A) The mean concentration of cytokines

produced in high-ABA subjects who were administered Imprime with

(n = 7) and without premedications (n = 5) are plotted. (B) Representative

example of one subject who received no premedication showing in-

creasing ABA levels and cytokine production with repeated dosing of

Imprime and one premedicated subject who had minimal cytokine re-

sponse to all three doses in spite of a spike in ABA levels. *p , 0.05,

***p , 0.001.

8 ANTI–b-GLUCAN Ab–DEPENDENT PHARMACODYNAMICS OF IMPRIME PGG

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 9: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

subclass of the IgG ABA that could preclude Imprime–IgGcomplex from interacting with the FcgRIIA receptor, are alsoimportant considerations for Imprime’s activity.It is conceivable that naturally occurring Abs cross-reacting to

Imprime or shared epitopes among structurally related b-glucansfrom different pathogens exist (25–27). ABA likely exists ascomponent of the natural Ab repertoire or as a reflection of ahumoral response to environmental Ags. The kinetics of in-crease in the IgG ABA levels (7 d postdose) and increased geneexpression of CD138 are suggestive of secondary activation ofpre-existing Ag-specific memory B cells or even release of CD138

positive plasma cells. Considering that the subjects showed noincrease in the IgM ABA levels, it is likely that the increase inABA levels is not a consequence of a primary de novo immuno-genicity response (Fig. 7). The cellular and molecular mech-anisms behind activation of the memory B cells remains tobe investigated, but we have shown that complement receptor 2(CR2) plays a vital role in binding of Imprime to B cells (28). AsCR2 is the coreceptor known to lower the threshold of BCR sig-naling, it is possible that Imprime is activating the memory B cellsthrough the bridging of b-glucan–specific BCR and CR2 (29–32).The subject-to-subject variability in this ABA increase could be

Table II. AE in healthy volunteers administered Imprime without premedications

Red indicates low-ABA, black indicates mid-ABA, and blue indicates high-ABA.{Delayed response.

The Journal of Immunology 9

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 10: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

due to multitude of reasons that affect a secondary recall response,including the clonal frequency or BCR affinity of the memoryB cells. Thus, although repeated dosing of Imprime could po-tentially be a strategy to overcome unresponsiveness to Imprimein patients with low baseline levels of ABA, it may not giveconsistent results. Instead, our previous work has shown that theadministration of IVIG to low-ABA patients can reliably increasethe ABA concentration to a level that restores Imprime-mediatedinnate pharmacodynamic responses (24).Understanding the IPD changes elicited by Imprime provided

the context for the infusion reactions observed with Imprime.The characteristics of the infusion reactions were suggestive ofmanifestations of complement activation-related pseudoallergy(33–35). The dynamics of the longitudinal changes in the ABAlevels also delineated the progression of infusion reactions acrossmultiple doses of Imprime. Increasing ABA levels were associatedwith heightened IPD responses to the second/third doses ofImprime (Fig. 7) as well as increased incidences of grade 2 and 3infusion reactions (Fig. 9). It is important to note that even thoughinfusion reactions were more common among subjects with

higher ABA levels, high ABA levels alone did not necessarilypredict the occurrence of a reaction. Further characterization ofthe infusion reaction, especially the type I hypersensitivity as-pect, is warranted. Even though our ex vivo studies have alreadyshown healthy volunteers to have a minimal IgE titer to Imprime(N. Bose and R.M. Walsh, unpublished observations), it wouldbe interesting to assess longitudinal change in IgE and tryptaselevels as markers of possible anaphylaxis.Premedication or concomitant treatment with corticoste-

roids is commonly employed for managing some of the clinicalimmunotherapy-associated infusion reactions or autoimmune sideeffects of several classes of anticancer or immune agonistic agents.Corticosteroids were recently shown to significantly affect theefficacy of checkpoint inhibitors in both preclinical and clinicalstudies (36, 37). It is important to understand the effect of corti-costeroid premedications on the pharmacodynamics of Imprime,as they have been the agent of choice to manage the infusionreactions. When used as premedications, corticosteroids substan-tially inhibited Imprime-mediated cytokine/chemokine produc-tion. Because cytokines/chemokines are paramount to mountingand coordinating an effective immune response, it is safe to as-sume that corticosteroids could hinder Imprime activity (Fig. 8).As an alternative to corticosteroids, several strategies have beenemployed to manage Imprime-related infusion reaction. In thistrial, subjects who experienced a grade 3 or greater (CTCAE)infusion reaction or were unable to complete dosing during weeks1 and/or 2 were, at the discretion of the investigator, allowed tobe rechallenged in the subsequent week with double the lengthof the standard infusion time. In our ongoing cancer studies(NCT02981303), an informal trial of steroid-sparing premed-ication regimen containing antihistamine (diphenhydramine),antiemetic (ondansetron), nonsteroidal anti-inflammatory drug(ketorolac), and antipyretic/analgesic agents (acetaminophen)is recommended before commencing dosing of study medicationson each treatment day and, as needed, postdose. To date, thispremedication regimen appears to be relatively effective in both

FIGURE 9. Imprime induces mild-to-moderate infusion-related re-

sponses. The maximum grade of any component symptom of the infusion

reaction associated with any dose of Imprime was plotted against each of

the predose ABA category.

Table III. AE in healthy volunteers administered Imprime with premedications

Red indicates low-ABA, black indicates mid-ABA, and blue indicates high-ABA.

10 ANTI–b-GLUCAN Ab–DEPENDENT PHARMACODYNAMICS OF IMPRIME PGG

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 11: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

managing the symptoms as well as preserving the Imprime-induced cytokine response in the treated subjects (24). Thesedata suggest that the use of steroids as supportive premedicationmay compromise the potential benefit of some immune-oncologyagents and, again, highlight the need for the medical communityto explore nonsteroid alternatives wherever possible.In summary, the data in this report provide in vivo evidence that

Imprime’s pharmacological activity is dependent on its binding tonaturally occurring ABA. The data presented demonstrate thatthere is a threshold of IgG ABA necessary for activation of theinnate immune responses. These data suggest that although theacute pharmacodynamic responses to Imprime-based immuno-therapy are predicated on ABA levels in individual healthy sub-jects, it is also likely that the high-ABA subjects will be moreprone to infusion-related reactions. It is imperative that a moreprudent use of corticosteroids be considered for Imprime-basedtherapy wherever possible. Ongoing phase 2 studies will providean opportunity to understand the correlation of these pharmaco-dynamic responses to clinical responses in cancer patients. Fur-thermore, exploration of this threshold in the context of otherfactors, including the subclass of the IgG, the concentration ofIgM ABA, the combination agent, the patient population, and theunderlying target disease, will be essential. This study provides astrong rationale for selecting patients based on their pretreatmentABA levels as a means to enrich for patients most likely to elicit aresponse to Imprime-based immunotherapy.

DisclosuresN.B., N.R.O., X.Q., B.H., R.M.W., A.S.H.C., M.T.U., and J.R.G. are

employed by and own stock/stock options in Biothera Pharmaceuticals,

Inc. B.T.R., J.R.L., A.B.J., K.B.G., T.O.K., L.R.J., K.E.E., R.B.F., S.M.L.,

K.A.F., and R.D.H. were Biothera employees at the time this study was per-

formed. M.A.M. is an employee of Prism Research, a contract research or-

ganization that received contractual remuneration for performance of the

clinical procedures of this study.

References1. Li, B., D. J. Allendorf, R. Hansen, J. Marroquin, D. E. Cramer, C. L. Harris, and

J. Yan. 2007. Combined yeast beta-glucan and antitumor monoclonal antibodytherapy requires C5a-mediated neutrophil chemotaxis via regulation of decay-accelerating factor CD55. Cancer Res. 67: 7421–7430.

2. Zhong, W., R. Hansen, B. Li, Y. Cai, C. Salvador, G. D. Moore, and J. Yan. 2009.Effect of yeast-derived beta-glucan in conjunction with bevacizumab for thetreatment of human lung adenocarcinoma in subcutaneous and orthotopic xe-nograft models. J. Immunother. 32: 703–712.

3. Salvador, C., B. Li, R. Hansen, D. E. Cramer, M. Kong, and J. Yan. 2008. Yeast-derived beta-glucan augments the therapeutic efficacy mediated by anti-vascularendothelial growth factor monoclonal antibody in human carcinoma xenograftmodels. Clin. Cancer Res. 14: 1239–1247.

4. Engel-Riedel, W., J. Lowe, P. Mattson, J. Richard Trout, R. D. Huhn, M. Gargano,M. L. Patchen, R. Walsh, M. M. Trinh, M. Dupuis, and F. Schneller. 2018. Arandomized, controlled trial evaluating the efficacy and safety of BTH1677 incombination with bevacizumab, carboplatin, and paclitaxel in first-line treatment ofadvanced non-small cell lung cancer. J. Immunother. Cancer 6: 16.

5. Zent, C. S., T. G. Call, D. A. Bowen, M. J. Conte, B. R. LaPlant, T. E. Witzig,S. M. Ansell, and G. J. Weiner. 2015. Early treatment of high risk chroniclymphocytic leukemia with alemtuzumab, rituximab and poly-(1-6)-beta-glucotriosyl-(1-3)- beta-glucopyranose beta-glucan is well tolerated and ach-ieves high complete remission rates. Leuk. Lymphoma 56: 2373–2378.

6. Zent, C. S., T. G. Call, T. D. Shanafelt, R. C. Tschumper, D. F. Jelinek,D. A. Bowen, C. R. Secreto, B. R. Laplant, B. F. Kabat, and N. E. Kay. 2008.Early treatment of high-risk chronic lymphocytic leukemia with alemtuzumaband rituximab. Cancer 113: 2110–2118.

7. Qi, C., Y. Cai, L. Gunn, C. Ding, B. Li, G. Kloecker, K. Qian, J. Vasilakos,S. Saijo, Y. Iwakura, et al. 2011. Differential pathways regulating innate andadaptive antitumor immune responses by particulate and soluble yeast-derivedb-glucans. Blood 117: 6825–6836.

8. Li, B., D. J. Allendorf, R. Hansen, J. Marroquin, C. Ding, D. E. Cramer, andJ. Yan. 2006. Yeast beta-glucan amplifies phagocyte killing of iC3b-opsonizedtumor cells via complement receptor 3-Syk-phosphatidylinositol 3-kinase path-way. J. Immunol. 177: 1661–1669.

9. Fulton, R. B., S. M. Leonardo, A. B. Jonas, K. A. Fraser, A. S. Chan,N. R. Ottoson, M. E. Danielson, N. Bose, J. R. Graff, and K. B. Gorden. 2016.

Imprime PGG, a b-glucan PAMP (pathogen-associated molecular pattern), ef-fectively elicits in vivo maturation of antigen presenting cells in mice and hu-mans, suggesting potential synergy with checkpoint inhibitor therapy (LB-089).Cancer Res. 76(14, Supplement): Abstract nr LB-089.

10. Bose, N., A. S. Chan, F. Guerrero, C. M. Maristany, X. Qiu, R. M. Walsh,K. E. Ertelt, A. B. Jonas, K. B. Gorden, C. M. Dudney, et al. 2013. Binding ofsoluble yeast b-Glucan to human neutrophils and monocytes is complement-dependent. Front. Immunol. 4: 230.

11. Chan, A. S., A. B. Jonas, X. Qiu, N. R. Ottoson, R. M. Walsh, K. B. Gorden,B. Harrison, P. J. Maimonis, S. M. Leonardo, K. E. Ertelt, et al. 2016. ImprimePGG-mediated anti-cancer immune activation requires immune complex for-mation. PLoS One 11: e0165909.

12. Thomas, M., P. Sadjadian, J. Kollmeier, J. Lowe, P. Mattson, J. R. Trout,M. Gargano, M. L. Patchen, R. Walsh, M. Beliveau, et al. 2017. A randomized,open-label, multicenter, phase II study evaluating the efficacy and safety ofBTH1677 (1,3-1,6 beta glucan; Imprime PGG) in combination with cetuximaband chemotherapy in patients with advanced non-small cell lung cancer. Invest.New Drugs 35: 345–358.

13. U. S. Department of Health and Human Services; National Institutes of Health;National Cancer Institute. 2009. Common Terminology Criteria for AdverseEvents (CTCAE). National Cancer Institute, Bethesda, MD.

14. Coutinho, A. E., and K. E. Chapman. 2011. The anti-inflammatory and immu-nosuppressive effects of glucocorticoids, recent developments and mechanisticinsights. Mol. Cell. Endocrinol. 335: 2–13.

15. Shoenfeld, Y., Y. Gurewich, L. A. Gallant, and J. Pinkhas. 1981. Prednisone-induced leukocytosis. Influence of dosage, method and duration of administra-tion on the degree of leukocytosis. Am. J. Med. 71: 773–778.

16. Griffith, J. W., C. L. Sokol, and A. D. Luster. 2014. Chemokines and chemokinereceptors: positioning cells for host defense and immunity. Annu. Rev. Immunol.32: 659–702.

17. Charo, I. F., and R. M. Ransohoff. 2006. The many roles of chemokines andchemokine receptors in inflammation. N. Engl. J. Med. 354: 610–621.

18. Cramer, D. E., S. Wagner, B. Li, J. Liu, R. Hansen, R. Reca, W. Wu, E. Z. Surma,D. A. Laber, M. Z. Ratajczak, and J. Yan. 2008. Mobilization of hematopoieticprogenitor cells by yeast-derived beta-glucan requires activation of matrixmetalloproteinase-9. Stem Cells 26: 1231–1240.

19. Ito, K., Y. Masuda, Y. Yamasaki, Y. Yokota, and H. Nanba. 2009. Maitake beta-glucan enhances granulopoiesis and mobilization of granulocytes by increasingG-CSF production and modulating CXCR4/SDF-1 expression. Int. Immuno-pharmacol. 9: 1189–1196.

20. Engel, A. L., G. E. Holt, and H. Lu. 2011. The pharmacokinetics of Toll-likereceptor agonists and the impact on the immune system. Expert Rev. Clin.Pharmacol. 4: 275–289.

21. Tisoncik, J. R., M. J. Korth, C. P. Simmons, J. Farrar, T. R. Martin, and M. G. Katze.2012. Into the eye of the cytokine storm. Microbiol. Mol. Biol. Rev. 76: 16–32.

22. Larkin, B., V. Ilyukha, M. Sorokin, A. Buzdin, E. Vannier, and A. Poltorak.2017. Cutting edge: activation of STING in T cells induces type I IFN responsesand cell death. J. Immunol. 199: 397–402.

23. Uhlik, M. T., B. Harrison, K. B. Gorden, S. M. Leonardo, R. M. Walsh,K. E. Ertelt, M. A. Gargano, M. Chisamore, J. Lowe, B. Osterwalder, et al. 2018.Imprime PGG, a soluble yeast b-glucan PAMP, in combination with Pem-brolizumab induces infiltration and activation of both innate and adaptive im-mune cells within tumor sites in melanoma and triple-negative breast cancer(TNBC) patients (LB-129). Cancer Res. 78(13, Supplement).

24. Ottoson, N. R., A. B. Jonas, A. S. Chan, X. Qiu, B. Rathman, R. M. Walsh,B. Harrison, M. E. Danielson, K. S. Michel, M. Finley, et al. 2018. Immuno-pharmacodynamic (IPD) responses of Imprime PGG combined with Pem-brolizumab in chemotherapy-resistant metastatic triple negative breast cancer (TNBC)subjects in a Phase 2 trial: analyses of Stage 1 patients (P708). J. Immunother.Cancer 6(Suppl. 2): 117.

25. Masuzawa, S. Y. M., K. Ishibashi, N. Saito, M. Akashi, N. Yoshikawa, T. Suzuki,S. Nameda, N. N. Miura, Y. Adachi, and N. Ohno. 2003. Solubilized candida cellwall b-glucan, CSBG, is an epitope of natural human antibody. Drug Dev. Res.58: 179–189.

26. Chiani, P., C. Bromuro, A. Cassone, and A. Torosantucci. 2009. Anti-beta-glucanantibodies in healthy human subjects. Vaccine 27: 513–519.

27. Ishibashi, K., M. Morita, M. Motoi, Y. Liu, N. N. Miura, Y. Adachi, andN. Ohno. 2013. Analysis of the titer and reactivity of antibody/ies against fungalcell wall b-glucans in human sera. Int. J. Med. Mushrooms 15: 115–126.

28. Bose, N., A. S. Chan, A. B. Jonas, X. Qiu, and W. Grossman. 2013. Recognitionof soluble yeast b-1,3/1,6 glucan by complement receptor 1 is essential for it tobind complement receptor 3 on human neutrophils and monocytes, and com-plement receptor 2 on B cells (P3048). J. Immunol. 190(1, Supplement).

29. Thornton, B. P., V. Vetvicka, and G. D. Ross. 1996. Function of C3 in a humoralresponse: iC3b/C3dg bound to an immune complex generated with natural an-tibody and a primary antigen promotes antigen uptake and the expression ofco-stimulatory molecules by all B cells, but only stimulates immunoglobulinsynthesis by antigen-specific B cells. Clin. Exp. Immunol. 104: 531–537.

30. Dempsey, P. W., M. E. Allison, S. Akkaraju, C. C. Goodnow, and D. T. Fearon.1996. C3d of complement as a molecular adjuvant: bridging innate and acquiredimmunity. Science 271: 348–350.

31. Haas, K. M., and T. F. Tedder. 2005. Role of the CD19 and CD21/35 receptorcomplex in innate immunity, host defense and autoimmunity. Adv. Exp. Med.Biol. 560: 125–139.

32. Carter, R. H., and D. T. Fearon. 1992. CD19: lowering the threshold for antigenreceptor stimulation of B lymphocytes. Science 256: 105–107.

The Journal of Immunology 11

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 12: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

33. Doessegger, L., and M. L. Banholzer. 2015. Clinical development methodologyfor infusion-related reactions with monoclonal antibodies. Clin. Transl. Immu-nology 4: e39.

34. van der Kolk, L. E., A. J. Grillo-Lopez, J. W. Baars, C. E. Hack, and M. H. vanOers. 2001. Complement activation plays a key role in the side-effects of rit-uximab treatment. Br. J. Haematol. 115: 807–811.

35. Buysmann, S., C. E. Hack, F. N. van Diepen, J. Surachno, and I. J. ten Berge.1997. Administration of OKT3 as a two-hour infusion attenuates first-dose sideeffects. Transplantation 64: 1620–1623.

36. Giles, A. J., M. N. D. Hutchinson, H. M. Sonnemann, J. Jung, P. E. Fecci,

N. M. Ratnam, W. Zhang, H. Song, R. Bailey, D. Davis, et al. 2018.

Dexamethasone-induced immunosuppression: mechanisms and implications for

immunotherapy. J. Immunother. Cancer 6: 51.37. Faje, A. T., D. Lawrence, K. Flaherty, C. Freedman, R. Fadden, K. Rubin,

J. Cohen, and R. J. Sullivan. 2018. High-dose glucocorticoids for the treatment

of ipilimumab-induced hypophysitis is associated with reduced survival in pa-

tients with melanoma. Cancer 124: 3706–3714.

12 ANTI–b-GLUCAN Ab–DEPENDENT PHARMACODYNAMICS OF IMPRIME PGG

by Nandita B

ose on April 15, 2019

http://ww

w.jim

munol.org/

Dow

nloaded from

Page 13: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

Supplemental Fig. 1. Imprime induces cytopenia in high-ABA subjects early into the infusion. CBC with differentials were obtained at various time pints and a fold change over the pre-dose value was calculated for neutrophils (top panels), monocytes (middle panels), and lymphocytes (bottom panels). Each line represents an individual subject in each of the ABA groups.

Pre-d

ose

15 m

in 4 h 24 h

0.0

0.5

1.0

1.5

Lym

phoc

yte

Cou

nt

(Fol

d C

hang

e)

2537

3967

Pre-d

ose

15 m

in 4 h 24 h

0.0

0.5

1.0

1.5

4347

5768

Pre-d

ose

15 m

in 4 h 24 h

0.0

0.5

1.0

1.5

5570

0.0

0.5

1.0

1.5P

MN

Cou

nt

(Fol

d C

hang

e)

0

1

2

3

4

0

1

2

3

4

0.0

0.5

1.0

1.5

Mon

ocyt

e C

ount

(F

old

Cha

nge)

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

Low-ABA Mid-ABA High-ABA

Page 14: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

Supplemental Fig 2. Imprime-induced IPD changes with repeated dosing. Serological changes in ABA levels and IPD changes with repeated dosing of Imrime were measured using the methods previously described. (A) A representative example of a mid-ABA subject showing, a) increasing ABA levels with repeated dosing of Imprime and thereby enhanced IPD responses as measured by increased b) CIC levels, c) SC5b-9 levels, d) cytokine production, e) fold change in neutrophil counts, and f) gene expression changes. (B) A representative example of a high-ABA subject showing, a) increasing ABA levels with repeated dosing of Imprime and thereby enhanced IPD responses as measured by increased b) CIC levels, c) SC5b-9 levels, d) cytokine production, e) fold change in neutrophil counts, and f) gene expression changes.

CD83ICAM1CD274ITGAMFCGR1BFCGR3BC5aRA1FCGR2AFCGR3ACD86CXCL8CCL4CXCL2CCL3CCL2CXCL1CXCR4CXCR1CCR1CCR5CCL5IL1RNIL10IL1BIL1ATNFAMX2OAS2SDCBPCD38

-2 0 2 4

CD83ICAM1CD274ITGAMFCGR1BFCGR3BC5aRA1FCGR2AFCGR3ACD86CXCL8CCL4CXCL2CCL3CCL2CXCL1CXCR4CXCR1CCR1CCR5CCL5IL1RNIL10IL1BIL1ATNFAMX2OAS2SDCBPCD38

-5 0 5

Pre-do

se EOI24

h

Pre-do

se EOI24

h

Pre-do

se EOI24

h16

8 h0

20

40

60

80

100

IgG

ABA

(ug/

ml)

Pre-do

se EOI

24h E

OI

Pre-do

se EOI

24h E

OI

Pre-do

se EOI

24h E

OI

168h

EOI

0

2

4

6

8

CIC

(Fol

d C

hang

e)

Pre-do

se EOI

Pre-do

se EOI

Pre-do

se EOI0

20

40

60

80

SC5b

-9 (F

old

Cha

nge)

Dose 1Dose 2

Dose 3

Dose 1 Dose 2 Dose 3 Dose 1 Dose 2 Dose 3 Dose 1 Dose 2 Dose 3

Pre-d

ose

15-6

0 min 4 h 24

h

Pre-d

ose

15-6

0 min 4 h 24

h

Pre-d

ose

15-6

0 min 4 h 24

h16

8 h0

1

2

3

PM

N C

ount

(Fol

d C

hang

e)

Dose 1 Dose 2 Dose 3

Cytokine

Type I IFNB cell

Innate Immune

Chemokine0

1

10

100IL-6

IL-8

IP-10

MCP-1

MIGMIP-1!

MIP-1β

TNF-!

IL-10

Dose 1Dose 2Dose 3

a) b) c)

d) e) f)

A.

Pre-do

se EOI24

h

Pre-do

se EOI24

h

Pre-do

se EOI24

h16

8 h0

50

100

150

IgG

ABA

(ug/

ml)

Pre-do

se EOI

24h E

OI

Pre-do

se EOI

24h E

OI

Pre-do

se EOI

24h E

OI

168h

EOI

0

1

2

3

4

5

CIC

(Fol

d C

hang

e)

Pre-do

se EOI

Pre-do

se EOI

Pre-do

se EOI0

5

10

15

SC5b

-9 (F

old

Cha

nge)

Dose 1 Dose 2 Dose 3 Dose 1 Dose 2 Dose 3 Dose 1 Dose 2 Dose 3

Pre-d

ose

15-6

0 min 4 h 24

h

Pre-d

ose

15-6

0 min 4 h 24

h

Pre-d

ose

15-6

0 min 4 h 24

h16

8 h0

2

4

6

PM

N C

ount

(Fol

d C

hang

e) Dose 1 Dose 2 Dose 3Dose 1

Dose 2Dose 3

0

1

10

100

1000IL-6

IL-8

IP-10

MCP-1

MIGMIP-1!

MIP-1β

TNF-!

IL-10

Dose 1Dose 2Dose 3

a) b) c)

d) e) f)

B.

Cytokine

Type I IFNB cell

Innate Immune

Chemokine

Page 15: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

Supplemental Fig 3. Dexamethasone inhibits Imprime-induced cytokine production ex vivo. Whole blood from high ABA individuals were treated ex vivo with vehicle alone or Imprime at 25 µg/mL in the presence of dexamethasone (0.2 and 2 µM) over 24 hours. Plasma was collected by centrifugation and assayed by Luminex X-MAP multiplex platform. Representative data shown from n = 2 experiments using different donors.

IL-1β

IL-1 R

AIL-

6IL-

8

MCP-1

MIP-1α

MIP-1β

RANTES0

1000

2000

3000

1000020000300004000050000

VehicleImp (25)Imp (25)+ Dex (0.2 uM)Imp (25)+ Dex (2 uM)

Cyt

okin

e (p

g/m

L)

Page 16: Immune Pharmacodynamic Responses of the Novel Cancer ... · Imprime PGG (Imprime) is an i.v. administered, yeast b-1,3/1,6 glucan in clinical development with checkpoint inhibitors

Supplemental Fig 4. Imprime-induced IPD changes in pre-medicated subjects. Serological changes in ABA levels and IPD changes were measured using the methods previously described. Shown here are the maximum fold changes at ~EOI for (A) IgG ABA, (B) CIC, (C) SC5b-9, and (D) CBC in low-, mid-, and high-ABA groups. Each point represents an individual subject in each of the ABA groups.

0

5

10100

150

200

SC

5b-9

(Fol

d C

hang

e)

0.0

0.5

1.0

IgG

AB

A (F

old

Cha

nge)

0.82

0.580.47

Low-ABA Mid-ABA High-ABA0

1

2

3

4

CIC

(Fol

d C

hang

e)

Low-ABA Mid-ABA High-ABA

1.24 2.35

1.88

A) B)

1.92 3.14128.2

C)

Low-ABA Mid-ABA High-ABA

0

2

4

6

CB

C (F

old

Cha

nge)

Low-ABAMid-ABA

High-ABA

Low-ABAMid-ABA

High-ABA

Low-ABAMid-ABA

High-ABA

Neutrophils Monocytes Lymphocytes

2.27

5.12

2.78

1.67

2.90

1.42 1.71 1.921.80

D)