Alternative complement pathway hemolytic assays reveal incomplete complement blockade in patients treated with eculizumab

Accepted Manuscript Alternative complement pathway hemolytic assays reveal incomplete complement blockade in patients treated with eculizumab

Bénédicte Puissant-Lubrano, Sylvain Puissochet, Nicolas CongyJolivet, Dominique Chauveau, Stéphane Decramer, Arnaud Garnier, Antoine Huart, Nassim Kamar, David Ribes, Antoine Blancher PII: DOI: Reference:

S1521-6616(17)30197-3 doi: 10.1016/j.clim.2017.06.007 YCLIM 7874

To appear in:

Clinical Immunology

Received date: Revised date: Accepted date:

17 March 2017 18 May 2017 20 June 2017

Please cite this article as: Bénédicte Puissant-Lubrano, Sylvain Puissochet, Nicolas Congy-Jolivet, Dominique Chauveau, Stéphane Decramer, Arnaud Garnier, Antoine Huart, Nassim Kamar, David Ribes, Antoine Blancher , Alternative complement pathway hemolytic assays reveal incomplete complement blockade in patients treated with eculizumab. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Yclim(2017), doi: 10.1016/j.clim.2017.06.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Alternative complement pathway hemolytic assays reveal incomplete complement blockade in patients treated with eculizumab

Bénédicte Puissant-Lubranoa,b, Sylvain Puissochetb, Nicolas Congy-Joliveta,b, Dominique Chauveauc , Stéphane Decramerd, Arnaud Garnierd, Antoine Huartc , Nassim Kamarc,e, David

T

Ribes c , Antoine Blanchera,b

SC RI P

a- Laboratoire d’Immunologie, CHU de Toulouse, Hôpital Rangueil, Toulouse, France. b- Laboratoire d’Immunogénétique Moléculaire, Université Paul Sabatier, Toulouse 3, Toulouse, France.

NU

c- Department of Nephrology and Organ Transplantation, CHU Rangueil, Toulouse, France

MA

d- Department of Pediatric Nephrology, CHU Purpan, Toulouse, INSERM 1048, Institute of Cardiovascular and Metabolic Diseases, Toulouse. Université de Toulouse III Paul Sabatier,

ED

Toulouse, France

e- INSERM U1043, IFR–BMT, CHU Purpan, Toulouse, Université Paul Sabatier, Toulouse,

PT

France [email protected]

CE

[email protected] [email protected]

AC

[email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

ACCEPTED MANUSCRIPT *Corresponding author and request for reprints to: Dr. Bénédicte Puissant-Lubrano Laboratoire d’Immunologie du CHU de Toulouse, Hôpital Rangueil TSA 50032 31059 Toulouse Cedex 9 France

SC RI P

Fax: 33 (0)5 61 32 34 24

T

Tel: 33 (0)5 61 32 34 32

AC

CE

PT

ED

MA

NU

E-mail: [email protected]

ACCEPTED MANUSCRIPT Abstract

Eculizumab is a monoclonal anti-C5 antibody used in the treatment of atypical hemolytic uremic syndrome (aHUS). We monitored complement inhibition in 16 eculizumab-treated patients suffering from HUS or transplant rejection (not aHUS patients). Blood samples were obtained one to four weeks after the last eculizumab injection.

T

We observed that eculizumab efficiently blocked the terminal pathway (TP) through

SC RI P

classical pathway (CP) activation measured by kinetic hemolytic assay (HA) (<10%) but incompletely blocked the TP through alternative pathway (AP) activation measured by rabbit (APH50>23%) or chicken erythrocytes HA (AP100 >15%). Conversely, functional ELISA revealed a complete blockade of TP through AP activation in all patients (<10%). C5a and sC5b9 levels were not correlated with residual APH50 or AP100.

NU

Similar results were obtained after in vitro addition of increasing amounts of eculizumab to a control serum (in vitro APH50 >60% and AP100 >20%). We also showed that ELISA was

MA

less sensitive than HA. Keywords

Complement classical pathway; complement alternative pathway; anti-C5; hemolytic assay;

PT

ED

ELISA; assay sensitivity; drug monitoring

Highlights

Eculizumab incompletely inhibited terminal pathway through alternative pathway activation

CE

-

Residual terminal pathway activity is observed in the majority of treated patients

-

This residual activity is observed only with hemolytic assays

-

ELISA in less sensitive than hemolytic assay to measure complement activity

AC

-

ACCEPTED MANUSCRIPT 1. Introduction

Eculizumab is an anti-C5 monoclonal antibody. The binding of this therapeutic MAb to the complement fraction C5 inhibits the activation of the latter by C5 convertases blocking the formation of C5b and of the terminal complement complex C5b-9. Eculizumab is indicated in the treatment of paroxysmal nocturnal hemoglobinuria (PNH) and atypical hemolytic uremic syndrome (aHUS). HUS is a life threatening thrombotic microangiopathy that

T

associates anemia, renal injury and thrombopenia. In its typical form it is associated with a

SC RI P

Shiga-toxin-producing E. coli (STEC-HUS). In its atypical form (aHUS), it is mainly due to deficiency of regulatory factors of the complement (Factor H, Factor I, MCP…) but also defective cobalamine deficiencies [1].

Routine evaluation of complement activation includes C3 and C4 quantification as well as

NU

assessment of the classical pathway (CP) activity. The alternative pathway (AP) activity is more rarely evaluated in clinical laboratories. However, the exploration of alternative pathway is of crucial importance in case of aHUS and PNH because both pathologies are

MA

caused by AP dysregulation. Recent studies have evaluated the use of complement activity to monitor patients treated with eculizumab. The classical pathway activity was associated with disease activity in PNH patients [2]. Two other studies showed that the quantification of

ED

classical pathway and alternative pathway (by ELISA) can help adapting the schedule of eculizumab administration in aHUS patient [3, 4]. Noris et al. proposed an assay that

PT

evaluates the C5b-9 deposit on endothelial cells [5] which seems to better parallel the clinical evolution of aHUS patients than CH50. However, this assay is specialized and cannot be performed routinely. Few studies have measured the AP activity to monitor

CE

eculizumab treatment, and those who did mainly used ELISA (in patients suffering from

AC

HUS [4] or PNH [6]).

In the present study, we explored the inhibitory capacity of eculizumab on the complement terminal pathway (TP) using different techniques: hemolytic assays of sensitized sheep red blood cells (RBCs) for CP and rabbit and chicken erythrocytes for AP, as well as functional ELISA. We quantified classical and alternative pathway activities on sera from treated patients and after in vitro addition of eculizumab on control sera. We showed that ELISA measuring complement activity had a lower sensitivity than hemolytic assays and was not able to detect a residual activity of TP through AP activation. In addition, we reported a residual TP activity (when AP was measured with HA) in 14 out of the 16 patients under eculizumab treatment explored here (in 7 of 9 aHUS patients).

ACCEPTED MANUSCRIPT

2. Patients, Materials and Methods 2.1.

Patients

Sixteen patients treated with eculizumab were studied (Table 1). These patients suffered from

STEC-HUS

(N=2),

atypical

HUS

(N=9)

or

from

renal

failure

(one

membranoproliferative glomerulonephritis with C3Nef, three patients treated for renal

T

transplant rejection and one PNH). Complement results before the initiation of eculizumab

SC RI P

treatment (C3, TH50 and APH50) were obtained from the medical file of the patients (the quantifications were performed in our laboratory). Serum samples collected before eculizumab initiation were kept frozen only for three patients (because we were informed that they would be treated by eculizumab). Therefore, C5a and sC5b9 concentrations before eculizumab initiation were measured only in these three patients because these

NU

parameters were not part of our routine exploration (see below).

Thirty samples (one to three samples per patient) were obtained just prior to a new infusion of eculizumab and between 7 days and 28 days after the previous injection (mean 14 days).

MA

Samples from these patients were addressed to our laboratory to monitor efficiency of their treatment by quantification of complement activity. Complement activity assessment by kinetic hemolytic assays

ED

2.2.

The kinetic hemolytic assays for quantification of classical and alternative pathways were

PT

adapted from the Mayer method and from Joiner, respectively [7, 8]. Briefly, we used kinetic hemolytic assays in microplate. In presence of samples (sera from patients or control sera with eculizumab added in vitro), the hemolysis of sensitized SRBCs

CE

(for CP, TH50) or rabbit erythrocytes (for AP, APH50) is assessed by a continuous measure of the optical density (O.D.) at 540 nm by means of spectrophotometer microplate reader reader/dispenser,

AC

(iEMS

ThermoFisher).

A

software

developed

by

Kirial

International/Laraspiral calculates the time (in seconds) to reach hemolysis of 50% of erythrocytes (TH50). The inverse of the TH50 is linearly correlated to the complement activity. Complement activity is deduced from the calibration range and is expressed as a percentage of the lysis produced with the calibrator. For details concerning the technique see Puissant-Lubrano et al. [9]. Alternative pathway was also measured by hemolysis of chicken erythrocytes in gel (Alternative pathway hemolytic complement kit, The Binding Site, Saint-Egrève, France). The chicken hemolytic test (AP100) was performed according to the manufacturer’s instructions. Note that this assay is validated for in vitro research use.

ACCEPTED MANUSCRIPT An incomplete inhibition (i.e. a residual activity) of TP of APH50 was considered when a decrease of O.D. was observed (i.e. the hemolytic curve was not flat). On the same way, an incomplete inhibition of TP of AP100 was considered when a lysis ring was observed in the gel. 2.3.

Complement activity assessment by ELISA

Classical and alternative complement activities were also assessed by ELISA (Wieslab®

T

Complement System, Alternative pathway and Classical Pathway, Diasorin, France). These

SC RI P

ELISA are functional techniques and are based on the quantification of the membrane attack complex (C5b-9) resulting from the in vitro activation of complement by the classical or the alternative pathway, depending on the activator deposited on the plate. ELISA were performed according to the manufacturer’ instructions.

Complement activity assessment after in vitro addition of eculizumab to a pool of

NU

2.4.

control sera

We performed a dose-response curve of inhibition of TP of classical and alternative

MA

pathways by eculizumab. Increasing amounts of eculizumab obtained from a vial residue (final concentrations from 0 (negative control) to 240 µg/mL) were added to a pool of control sera obtained from healthy blood donors. The initial complement activity of the pool was

ED

always >70% for CP and AP (four independent experiments). Classical and alternative complement activities were determined by hemolytic assays and ELISA. A fifth experiment

PT

was performed on TH50 and APH50. The supernatants were recovered, centrifuged and frozen until use.

Quantification of C5a and sC5b-9

CE

2.5.

C5a and sC5b-9 were quantified in the 30 serum samples from eculizumab-treated patients

AC

as well as in the supernatants obtained after a dose-response effect of eculizumab on TH50 and APH50 (see 2.4.). Quantification was also performed on i) sera obtained from three patients before eculizumab treatment (patients 6, 7 and 15) as well as sera obtained from two patients six weeks after eculizumab cessation (patients 3 and 5), ii) sera from 20 healthy blood donors from Etablissement Français du sang. Quantification was performed by ELISA (MicroVue C5a EIA and MicroVueComplement SC5b-9 Plus EIA, Quidel, distributed by Ingen, Chilly-Mazarin, France), according to the manufacturer’s instructions.

3. Results

ACCEPTED MANUSCRIPT 3.1.

Characteristics of the patients studied

Sixteen patients treated with eculizumab were studied. The patients suffered from STECHUS, atypical HUS or from chronic renal failure (patient 9 had a C3NeF, patients 1; 5; 10 has required a renal transplant) (Table 1). All but two patients had normal classical complement activity and normal C3 level before eculizumab treatment. The two remaining patients (patients 8 and 9) suffered from aHUS and had a C3 below 0.4 g/L. AP was quantified in seven patients before initiation of eculizumab treatment, only one patient

T

(patient 8) had low AP.

SC RI P

Twelve patients were in a stable phase of their disease and were treated by eculizumab for more than one month. The four remaining patients (1; 2; 4 and 7) were treated by eculizumab for one week and were not in a stable phase of their disease (at time of first sampling of the study). Eculizumab treatment significantly improved LDH (mean 1290 UI/L vs 486 UI/L, p=0.027 Wilcoxon test), hemoglobin (9.8 g/dL vs 11.3 g/dL, p=0.022) and

NU

platelets (124 375/mm 3 vs 183 078/mm 3, p=0.046) but not creatininemia (388 µmol/L vs 261 µmol/L). A decrease of creatininemia was observed in nine patients out of 16 (in five

3.2.

MA

aHUS patients and four non aHUS patients, table 1).

Complement activity in serum of patients treated with eculizumab

Alternative and classical complement activities were measured in the 30 sera obtained one

ED

to four weeks after the last eculizumab injection. The TP activity (measured either by kinetic TH50 or CP ELISA) was completely inhibited in all patients (<10%, Figure 1A). As for TP

PT

measured by kinetic APH50 (Figure 1B, open squares), all but one patient (patient 8) displayed a detectable activity (mean 42%, range 23-76% AP activity). Residual TP-APH50 activity was not correlated with the time since last eculizumab injection (data not shown).

CE

For TP measured by AP100, (Figure 1B, black triangles), all but two patients had a low but detectable activity (mean 33%, range 12-126% AP activity): a lysis ring can be observed.

AC

However, for all patients, the lysis was incomplete (intact erythrocytes in the lysis ring were observed under microscope). Therefore because the results presented are derived from the diameter of the lysis ring even if the lysis was incomplete, the results of AP100 are slightly overestimated. For patients 8 (see details above), patient 12 and the third sample of patient 4 a very slight lysis ring was observed and the results were negated. Patient 12 had C3 at 0.65 g/l and C4 at 0.27 g/L. She had an aHUS on a stable phase. All patients had AP<10% as measured by ELISA (Figure 1B, circles). The residual TP activity measured by AP100 was slightly higher in non-aHUS patients than in aHUS patients (50.2% s 22.4%, p=0.04 Mann-Whitney) while the difference was not statistically significant for APH50 activity (40 vs 45%, p=0.3).

ACCEPTED MANUSCRIPT

3.3.

Inhibition of the complement activity by eculizumab in vitro

Adding increasing amounts of eculizumab to a pool of human sera, we observed a concentration-dependent inhibition of TP measured by kinetic TH50 (CP) (Figure 1C). A complete inhibition of hemolysis was obtained for eculizumab concentration above 120 µg/mL (no significant decrease of OD was observed after 7.5 minutes corresponding to an activity <10%). The 50% inhibitory concentration calculated from four independent

T

experiments varies from 30 to 42 µg/mL. In control experiments, we used a commercially

SC RI P

available ELISA to confirm the inhibition of TP through CP activation by eculizumab. Eculizumab completely blocked TP measured by CP ELISA for concentration above 120 µg/mL and the 50% inhibition concentration was in a range similar to that observed with the hemolytic assay (24 to 38 µg/mL, Figure 1D).

By contrast, in assays with uncoated rabbit RBCs (APH50) we never observed a complete

NU

inhibition of hemolysis even in presence of 240 µg/mL of eculizumab (median APH50 activity 61%, range 56-67% in four independent experiments, Figure 1E). The 50% of maximum effect was calculated from four independent experiments between 37-50 µg/mL.

MA

By means of a commercially available kit based on hemolysis of chicken erythrocytes in gel (AP100, figure 1F), we also observed a residual TP activity even at high eculizumab concentrations (>120 µg/mL) (AP100 activity 15% and 24% in two independent

ED

experiments, Figure 1E). The 50% of maximum effect was between 34-59 µg/mL (two independent experiments). Contrary to the hemolytic tests, the ELISA revealed a complete

PT

inhibition of TP measured by AP for eculizumab concentrations higher than 60 µg/mL (Figure 1G). The 50% inhibitory concentration calculated from four independent

Quantification of C5a and sC5b9 in patients and in supernatants obtained after

AC

3.4.

CE

experiments was 25-39 µg/mL.

a dose-response effect of eculizumab on TH50 and APH50 Patient sera were obtained after the beginning of eculizumab treatment and were kept frozen until quantification of C5a and sC5b9. The sC5b9 was high (>1105 ng/mL) in six patients out of 16 (37.5%) and more precisely in 50% of aHUS patients in a stable phase of their disease and in none of the aHUS patients in the acute phase of the disease (Table 1). Only one patient (Patient 4) had high level of C5a (Table 1). Our results are in accordance with those previously reported [10, 11] except that the frequency of non-stabilized aHUS patients with normal sC5b9 level (100%) was higher than that previously reported by Noris et al. [5]. There was no significant association between sC5b9 concentration and APH50 or AP100 activity (Spearman p=0.75 and 0.58, respectively) or between C5a concentration

ACCEPTED MANUSCRIPT and APH50 or AP100 activity (Spearman p=0.58 and 0.075, respectively). In absence of eculizumab treatment (before initiation of treatment or after cessation of treatment), sC5b9 and C5a were higher than during eculizumab treatment in the five patients studied (Figure 2A). As controls, we also quantified sC5b9 and C5a after in vitro activation of complement by CP or AP (see materials and methods for details). As shown in figure 2B, the concentrations of sC5b9 and of C5a resulting from activation of alternative pathway ( APH50) decreased while

T

eculizumab concentration increased. The 50% inhibitory concentrations were 34 µg/mL for

SC RI P

C5a and 29 µg/mL for sC5b9, i.e. close to those observed with hemolytic assays. Comparable results were obtained after TH50 assay, taking into account the dilution of the serum in the assay.

Absence of complement-independent rabbit erythrocyte lysis

NU

3.5.

Because the residual hemolysis of rabbit erythrocytes obtained with samples in presence of eculizumab (i.e. serum of patients treated with eculizumab or control serum after in vitro

MA

addition of eculizumab) was discordant with ELISA results, we asked whether this lysis was complement-dependent. For that, we measured APH50 in presence of 10 control sera (from 10 different blood donors) after inactivating complement activity by: i) heating 30 min at

ED

56°C and ii) addition of EDTA to chelate Mg2 +. Rabbit RBCs hemolysis was fully inhibited (lack of OD decrease after 30 minutes) in presence of EDTA or after decomplementation at

PT

56°C. A persistent hemolysis of rabbit RBCs was observed for the same control sera (not inactivated) in presence of 240 µg/mL of eculizumab. Therefore, we concluded that rabbit erythrocytes lysis observed in presence of high concentration of eculizumab (240µg/ml)

Sensitivity of hemolytic assays and ELISA to measure alternative pathway

AC

3.6.

CE

was complement-dependent.

A different sensitivity of the assays could explain why a residual TP activity was observed when measuring AP by HA but not by ELISA in presence of high concentrations of eculizumab. Therefore, we quantified the alternative complement activity of dilutions of a pool of normal sera by means of i) APH50, ii) AP100 and iii) functional ELISA (Figure 3). Figure 3 shows that the two hemolytic assays were more sensitive than ELISA: the dilution corresponding to the theoretical activity of 34% of AP was clearly evidenced by rabbit and chicken erythrocytes hemolysis but not by ELISA. We concluded that contrary to ELISA, the hemolytic assay was capable to reveal the faint residual activation of complement TP by AP that was not fully blocked by eculizumab even at high concentrations (240µg/mL).

ACCEPTED MANUSCRIPT

4. Discussion In the present study, we showed that functional assays revealed discordant results concerning the blockade of complement TP through AP activation. The ELISA revealed a complete blockade while hemolytic assays of rabbit or chicken RBCs revealed a residual TP activity in presence of eculizumab. By contrast, the functional activity of complement

T

through CP activation was fully inhibited by eculizumab whatever the technique used

SC RI P

(kinetic hemolysic assay or ELISA). These results were observed in the serum of patients treated by eculizumab as well as after in vitro addition of eculizumab to a pool of sera from healthy blood donors.

Our observation of a complete inhibition of TP activity through CP activation is in

NU

accordance with previous studies performed in patients treated with eculizumab (with ELISA method [3, 12]; or with hemolytic assay [2]). In an in vitro assay, we have estimated the 50% inhibitory dose between 30 and 50 µl/mL of eculizumab, in accordance with results

MA

previously reported by Peffault de Latour et al [2]. The in vivo concentration of eculizumab was targeted between 50 and 100 µg/mL to fully inhibit TP activation measured by CP

ED

hemolytic assay [13].

In most previous reports, the eculizumab blockade of TP activity through alternative

PT

pathway activation in serum of patients was explored by ELISA (in HUS [3, 4] and in PNH [6]). In the latter studies, the authors reported a complete blockade of TP through AP activation in serum of patients treated by eculizumab. This is in accordance with the results

CE

we obtained by means of ELISA. To our knowledge, only one recent study explored the inhibitory effect of eculizumab on TP through AP activation by means of a rabbit erythrocyte

AC

hemolytic assay [14]. Harder et al. found that residual alternative pathway activity was observed after in vitro addition of eculizumab. In another recent study, Welhing et al. demonstrated that the TP of AP measured by chicken erythrocytes hemolytic assay was incompletely blocked in the serum of two patients treated by eculizumab [11]. These results are in agreement with our own results: i) we observed that addition of eculizumab to a pool of normal sera was unable to fully block the TP of AP in the rabbit or chicken hemolytic assays, ii) we observed that 14 out of 16 eculizumab-treated patients (mainly suffering from HUS) displayed a residual TP activity through alternative pathway activation as measured by hemolytic assays. Therefore, rabbit or chicken RBC hemolysis revealed a TP activity through alternative pathway activation in eculizumab-treated patients (the present study, [11, 14]) while sheep RBCS hemolytic assays revealed a lack of TP activity through CP

ACCEPTED MANUSCRIPT activation. Noteworthy the residual TP activity through alternative pathway activation was not revealed by ELISA which, in our hands, was less sensitive than AP hemolytic assays. We showed here that the lack of full inhibition of TP of AP by eculizumab seems to be a general feature in most of patients. We observed that sC5b9 and C5a were not correlated in aHUS patients, as previously reported [11] and that sC5b9 and C5a concentrations were not significantly associated with residual TP activity of APH50 or AP100. However, despite hemolytic tests revealed a residual TP activity through AP activation in sera obtained one to

T

four weeks after the last injection of eculizumab, the evolution of the disease was stabilized

SC RI P

in 12 out of the 16 patients studied here. Therefore, despite the eculizumab treatment showed a clinical efficacy, the blockade of alternative pathway was incomplete. As suggested by Harder et al., the addition of a second complement inhibitor could allow the full blockade of alternative complement pathway [14].

NU

The cause of the discrepancy between hemolytic assays and ELISA to measure the TP activity through AP activation in eculizumab-treated patients remains under discussion as well as the discrepancy between the residual TP activity revealed by AP hemolytic assays

MA

and the absence of residual TP activity through CP activation (whatever the technique used to reveal it). It was recently confirmed that eculizumab recognizes an epitope of C5, which is far from the cleavage site of C5 convertases [15]. The authors conclude that eculizumab

ED

acts most probably by inhibiting the binding of C5 to the C5 convertases rather than directly hindering the proteases C2a or Bb. Therefore, one cannot exclude that the binding of C5 to the classical C5 convertase (C4b-C2a-C3b) is more efficiently blocked by eculizumab than

PT

the binding of C5 to the alternative C5 convertase (C3b-Bb-C3b’). Unfortunately the two difference.

CE

natural C5 convertase are too unstable to allow a direct molecular demonstration of such a In addition, it has been shown both in mouse and human that C5 can be activated

AC

independently of convertases, notably by thrombin [16, 17]. Although this pathway causes a lower chicken erythrocyte lysis than that induced by convertases [17], it could participate to the residual TP observed under eculizumab treatment. Finally, two groups have recently reported concordant results on the mechanism of C5 activation by C5 convertases [14, 15]. These two studies highlighted the importance of the surface-bound C3b that prime C5 for cleavage by convertases. Under strong complement activation by AP, the great local concentration of C3b molecules compete with eculizumab for recruiting C5 in its primed conformation [14]. It has to be noted that the conditions of TH50 and APH50 in vitro assays differ, the TH50 usually being performed at low serum content contrary to APH50. Harder et al. have recently shown that a residual hemolysis persists in presence of eculizumab excess after forceful activation of CP [14]. Therefore,

ACCEPTED MANUSCRIPT they suggest a residual TP activity under eculizumab treatment both through AP and CP activation. However, we observed a residual TP activity in patients in a stable phase of their disease and to date, no link was established with the clinical outcome of the patients. More studies are necessary to establish a potential link between this residual TP activity and the clinical outcome of the patients.

T

In conclusion, our study revealed that ELISA was less sensitive than hemolytic assay. In

SC RI P

addition, we showed that the residual TP activity through AP activation one to four weeks after eculizumab injection seems to be a common feature in patients treated by eculizumab. By contrast, a sensitive hemolytic assay revealed no detectable residual TP activity through CP activation in any of the patients. In vitro inhibition of hemolytic assays by addition of eculizumab revealed a residual activity of TP through AP activation and a complete

NU

blockade of TP through CP activation.

MA

Remark: While we were writing this article, Harder et al. reported that alternative pathway measured by hemolysis of rabbit erythrocytes was not fully inhibited by eculizumab (in vitro

AC

CE

PT

ED

and in two PHN patients).

ACCEPTED MANUSCRIPT References

1.

Taylor CM, Machin S, Wigmore SJ, Goodship TH. Clinical practice guidelines for the management of atypical haemolytic uraemic syndrome in the United Kingdom. Br J Haematol 2010,148:37-47.

2.

Peffault de Latour R, Fremeaux-Bacchi V, Porcher R, et al. Assessing complement blockade in patients

with paroxysmal nocturnal hemoglobinuria receiving

Cugno M, Gualtierotti R, Possenti I, et al. Complement functional tests for

SC RI P

3.

T

eculizumab. Blood 2015,125:775-783.

monitoring eculizumab treatment in patients with atypical hemolytic uremic syndrome. J Thromb Haemost 2014,12:1440-1448. 4.

Volokhina EB, van de Kar NC, Bergseth G, et al. Sensitive, reliable and easyperformed laboratory monitoring of eculizumab therapy in atypical hemolytic uremic

5.

NU

syndrome. Clin Immunol 2015,160:237-243.

Noris M, Galbusera M, Gastoldi S, et al. Dynamics of complement activation in aHUS and how to monitor eculizumab therapy. Blood 2014,124:1715-1726. Hallstensen RF, Bergseth G, Foss S, et al. Eculizumab treatment during pregnancy

MA

6.

does not affect the complement system activity of the newborn. Immunobiology 2015,220:452-459.

Mayer MM. In : Experimental Immunochemistry. Eds EA Kabat and MM Mayer

ED

7.

1961,(Thomas Springfield, IL) Ch4. Joiner KA, Hawiger A, Gelfand JA. A study of optimal reaction conditions for an

PT

8.

assay of the human alternative complement pathway. Am J Clin Pathol 1983,79:659.

CE

72.

Puissant-Lubrano B, Fortenfant F, Winterton P, Blancher A. A microplate assay to measure classical and alternative complement activity. Clin Chem Lab Med 2017. Volokhina EB, Bergseth G, van de Kar NC, van den Heuvel LP, Mollnes TE.

AC

10.

Eculizumab treatment efficiently prevents C5 cleavage without C5a generation in vivo. Blood 2015,126:278-279. 11.

Wehling C, Amon O, Bommer M, et al. Monitoring of complement activation biomarkers and eculizumab in complement-mediated renal disorders. Clin Exp Immunol 2017,187:304-315.

12.

Gatault P, Brachet G, Ternant D, et al. Therapeutic drug monitoring of eculizumab: Rationale for an individualized dosing schedule. MAbs 2015,7:1205-1211.

13.

Legendre CM, Licht C, Muus P, et al. Terminal complement inhibitor eculizumab in atypical hemolytic-uremic syndrome. N Engl J Med 2013,368:2169-2181.

ACCEPTED MANUSCRIPT 14.

Harder MJ, Kuhn N, Schrezenmeier H, et al. Incomplete inhibition by eculizumab: mechanistic evidence for residual C5 activity during strong complement activation. Blood 2017,129:970-980.

15.

Jore MM, Johnson S, Sheppard D, et al. Structural basis for therapeutic inhibition of complement C5. Nat Struct Mol Biol 2016,23:378-386.

16.

Huber-Lang M, Sarma JV, Zetoune FS, et al. Generation of C5a in the absence of C3: a new complement activation pathway. Nat Med 2006,12:682-687.

T

Krisinger MJ, Goebeler V, Lu Z, et al. Thrombin generates previously unidentified

SC RI P

C5 products that support the terminal complement activation pathway. Blood

CE

PT

ED

MA

NU

2012,120:1717-1725.

AC

17.

ACCEPTED MANUSCRIPT Before ecul i zuma b Age (years)

Diagnosis

1

65

2

29

foca l s egmenta l hya l i nos i s rejecti on s us pi ci on a HUS

3

14

a HUS

4

1

a HUS

Mutation

Tr

Plasma exchange*

LDH UI/L

Hb g/dL

Creat. µmol/L

PQ /mm3

C3 g/L 0.721.51 0.8

Y

Y

135225 595

13-17.5 M 11-13.6 W

10

570

150 000450 000 62 000

MCP (p.Lys172Asn) CFI (p.Met138Val)

N

Y

1606

12.6

630

95 000

0.96

Y

N

317

9

680

213 000

0.97

recombination

N

N

6577

9

105

25 000

1.64

CFH and CFHR1 5

61

6

51

MPGN ki dney a l l ogra ft rejecti on a HUS

7

22

a HUS

8

45

a HUS

9

15

Fr

LDH

Hb

Creat. µmol/L

PQ /mm3

C5a ng/mL

W1 D21 W5 W1 D15 M9 M14 M20

1W 1W 2W 1W 1W 2W 3W 4W

689 764 882 974 644 534 437 191

11 7.9 13.7 10.2 8.6 11.2 12.7 11.9

471 369 448 699 708 61 60 69

64 000 68 000 74 000 293 000 241 000 251 000 252 000 228 000

8.9 11.2 18.6 15.2 20.8 10.2 12.7 13.2

T P

I R

sC5b9 ng/mL 2761105*** 394 410 985 526 482 653 797 515

5.6-26.0***

C S U W1

1W

1570

13.4

94

53 000

57.0

990

1W 3W 2W 3W 3W 2W 2W 1W 1W

673 262 150 290 291

4W

157 263 215 502 223 202

8.9 11.8 12.5 11.7 12.8 11.3 12.5 9.2 8.5 11.8 11.8 11.5

22 24 154 184 164 212 202 740 726 980 1091 101

463 000 270 000 267 000 285 000 268 000 176.000 176 000 125 000 106 000 194 000 159 000 266 000

12.2 8.4 9.8 10.1 18.0 13.8 7.2 5.8 9.6 7.9 13.3

499 471 559 576 1051 1123 581 913 1858 935 797

177 000

0.87

509

71 000

1.51

10.7

454

71 000

1.45

12.6

-

214 000

0.47

644

9.2

81

297000

0.35

AC

N

402

9.1

578

104000

0.80

M1.7

2W

614

12.2

664

norma l count

25.8

1498

N Y

N N

1455 310

7.6 12.7

204 132

65000 169 000

0.70

Y

728

8.6

489

33 000

1.28

2W 1W 2W 2W

282

N

M1.5 M1 M2 M5

215 223

10.7 13.4 12.8 12.7

31 138 137 79

254 000 215 000 203 000 235 000

6.4 11.5 8.2 13.5

747 907 907 1057

N

N

2470

10.3

227

150 000

1.02

M2.5

2W

314

8.4

304

166 000

20.9

5835

M3 M11 M1

2W 2W 1W

149 187 332

10.2 11.2 10.3

160 148 69

169 000 220 000 194 000

5.6 8.0 21.0

360 460 2250

CFHR5 (p.Arg356His) CFB (p.Lys565Glu) anti-CFH

Y

N

262

8.5

570

Y

N

645

9.4

N

N

306

Y

PT -

E C

Y

N

D E

M

N A

10

58

11 12

2 36

13

77

14

67

15

35

a HUS

none

N

N

172

9.3

184

209000

1.14

16

8

STEC-HUS

-

N

N

2867

7.9

404

35000

0.66

a HUS chroni c rena l fa i l ure, HNP

Time**

M3 M8 M2 M6 M7 M1 M1.5 W1 W3 2.8y 4.7y 3y

N nephroangioscl eros i s ki dney a l l ogra ft rejecti on STEC-HUS a HUS

After ecul i zuma b

CFH (p.Gly879Arg) none

ACCEPTED MANUSCRIPT Table 1. aHUS atypical hemolytic and uremic syndrome; Fr: frequency of eculizumab administration (every X week); MPGN membranoproliferative glomerulonephritis; TMA thrombotic microangiopathy; Tr Transplantation * before eculizumab ; ** Time since first eculizumab injection; *** reference values obtained on sera (mean +/- 2SD, see 2.5.)

T P

I R

C S U

N A

D E

T P E

A

C C

M

ACCEPTED MANUSCRIPT Figure legends

Figure 1. Effect of eculizumab on classical and alternative complement activity (A, B) Sera were obtained from 16 eculizumab-treated patients (one to three samples per patient). Classical (A) and alternative pathways (B) were measured by hemolytic assays (diamonds for CP and squares for AP) and ELISA (crosses for CP and circles for AP). (AP100, dark triangles). The results are expressed as % activity.

T

Alternative pathway was also measured with radial immunodiffusion of chicken RBCs

SC RI P

(C, D, E, F, G) Increasing amounts of eculizumab (final concentrations from 5 to 240 µg/mL) were added to pooled normal sera. Twelve concentrations of eculizumab, including a control without eculizumab were obtained. Functional activity of TP of AP and CP were quantified by hemolytic assays (C: TH50 with antibody-coated sheep RBCs, E: APH50 with uncoated rabbit RBCs; F: AP100 with chicken RBCs) and ELISA (D, G). The nonlinear regression

NU

based on Boltzmann equation (GraphPad® software) resulting from four independent experiments is shown. For each concentration of eculizumab, the mean of complement

MA

activity +/- SD is shown. Dashed lines represent the lower and the upper asymptotes from which is deduced the concentration of eculizumab allowing the 50% inhibition of the activity. For RID, one experiment representative of two is shown, nine concentrations of eculizumab

ED

are tested.

Figure 2. Effect of eculizumab on C5a and sC5b9

PT

A. C5a and sC5b9 were quantified in serum obtained from five patients after eculizumab treatment cessation (grey circles). The mean of the values obtained under eculizumab

CE

treatment (see Table 1) are shown (open circles). B. In vitro inhibition of C5a and C5b9 production by eculizumab. Increasing amounts of

AC

eculizumab (final concentrations from 0 to 240 µg/mL) were added to pooled normal sera. APH50 assay was performed. After 34 minutes of incubation at 37°C, the microplate was centrifuged and the supernatants were recovered for quantification of C5a and sC5b9 by ELISA. The figure shows the nonlinear regressions based on Boltzmann equation (GraphPad® software) for C5a (open squares) and sC5b9 (black squares), according to the concentration of eculizumab added.

ACCEPTED MANUSCRIPT Figure 3. Sensitivity of alternative complement activity measured with hemolytic assays or ELISA. The activation of complement by the alternative pathway was quantified in serial dilutions of a pool of normal sera by means of i) APH50 (open squares), ii) AP100 (black triangles) and iii) ELISA (open circles). Measured activities are represented versus the theoretical values. Linear regression is shown for each assay taking into account measurable activities (>10%).

AC

CE

PT

ED

MA

NU

SC RI P

T

This experiment is representative of three independent experiments.

AC

CE

PT

ED

MA

NU

SC RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC RI P

T

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC RI P

T

Figure 3