The Journal : Current Issue

Jan ter Meulen, Alexander B H Bakker, Edward N van = den=20 Brink, Gerrit J Weverling, Byron E E Martina, Bart L = Haagmans, Thijs=20 Kuiken, John de Kruif, Wolfgang Preiser, Willy Spaan, Hans R = Gelderblom, Jaap Goudsmit, Albert D M E Osterhaus

SARS coronavirus continues to cause sporadic cases = of=20 severe acute respiratory syndrome (SARS) in China. No active = or=20 passive immunoprophylaxis for disease induced by SARS = coronavirus is=20 available. We investigated prophylaxis of SARS coronavirus = infection=20 with a neutralising human monoclonal antibody in ferrets, = which can=20 be readily infected with the virus. Prophylactic = administration of=20 the monoclonal antibody at 10 mg/kg reduced replication of = SARS=20 coronavirus in the lungs of infected ferrets by 3=B73 logs = (95% CI=20 2=B76-4=B70 logs; p<0=B7001), completely prevented the = development of=20 SARS coronavirus-induced macroscopic lung pathology = (p=3D0=B7013), and=20 abolished shedding of virus in pharyngeal secretions. The = data=20 generated in this animal model show that administration of a = human=20 monoclonal antibody might offer a feasible and effective = prophylaxis=20 for the control of human SARS coronavirus infection.=20

Lancet 2004; 363: 2139-41=20

See Commentary=20

Severe acute respiratory syndrome (SARS) has emerged = as a=20 frequently fatal respiratory-tract infection caused by the = newly=20 identified SARS coronavirus. After the worldwide SARS = epidemic in=20 2002-2003, sporadic cases continue to arise in southern = China,=20 possibly because of human contact with the animal=20 reservoir.¹ Two recent cases of = laboratory-acquired SARS=20 coronavirus infections in China spread into the community = and=20 triggered extensive efforts in tracing and isolating = contacts of=20 patients to prevent a new epidemic. Means to control SARS=20 coronavirus infection through active or passive immunisation = are,=20 therefore, urgently needed.=20

Passive transfer of mouse immune serum has been shown = to=20 reduce pulmonary viral titres in mice infected with SARS=20 coronavirus.² Immunoprophylaxis of SARS = coronavirus=20 infection with human monoclonal antibodies might therefore = be a=20 viable strategy to control SARS.=20

We generated a human IgG1 monoclonal antibody, = CR3014,=20 reactive with whole inactivated SARS coronavirus, by = antibody phage=20 display technology screening a large naive antibody library. = Binding=20 of this antibody to the viral peplomers was visualised by = electron=20 microscopy with indirect two-step immunogold labelling. SARS = coronavirus from the supernatant of infected Vero cells was = adsorbed=20 to copper grids coated with carbon and Pioloform, which were = incubated with the monoclonal antibodies by floating on = droplets for=20 30 min at room temperature. Bound monoclonal antibodies were = detected by incubation on droplets of anti-human-IgG-gold-5 = nm=20 conjugates (British Biocell, Cardiff, UK). The grids were = negative=20 contrasted with 1% uranyl acetate and assessed with a ZEISS = EM 10 A=20 transmission electron microscope (figure 1A, 1B). CR3014 was = shown=20 to react with the cell-surface expressed spike glycoprotein = of SARS=20 coronavirus: HEK293T cells were transfected with the=20 plasmid-expressed, codon-optimised full-length glycoprotein = spike,=20 and stained with CR3014 and an R-phycoerythrin-labelled = second=20 antibody after 48 h using standard procedures for = FACS-analysis=20 (figure 1C). CR3014 was also shown to neutralise SARS = coronavirus in=20 vitro: triplicate wells of Vero cells were infected with 100 = TCID₅₀ (50% tissue culture infective dose) of = SARS=20 coronavirus strain HK-39849, and pre-incubated for 1 h at = 37=B0 C with=20 different concentrations of CR3014 or control antibody. The = presence=20 or absence of a cytopathic effect was scored in each well on = day 5=20 after infection, and the percentage of protected wells was = noted=20 (figure 1D).=20

Figure 1:Binding of CR3014 = monoclonal=20 antibody to viral peplomers of SARS coronavirus (A, B) = and to=20 HEK293T cells expressing glycoprotein spike of SARS=20 coronavirus (C), and in vitro neutralisation of SARS=20 coronavirus strainHK-39849 (D)

Incubation with CR3014 led to a dense = gold-label of the=20 outer peplomer region of SARS coronavirus strain = Frankfurt 1=20 (A), whereas a control human IgG1 monoclonal antibody = did not=20 induce any label (B). Bar indicates 100 nm. (C) Filled = histogram=3Dglycoprotein spike of SARS coronavirus, = strain=20 Frankfurt 1. Open histogram=3Dcontrol plasmid.=20

We tested the in-vivo potency of this antibody in = ferrets,=20 because these animals can be readily infected with SARS = coronavirus=20 by the intratracheal route, replicate the virus to high = titres in=20 the lung, and develop macroscopic and microscopic lung=20 lesions.³ Experimental infection of ferrets with = human=20 respiratory viruses, including influenza viruses, is common=20 practice, therefore procedures to work with these animals = under the=20 biocontainment conditions required for SARS coronavirus are = well=20 established. Approval for animal experiments was obtained = from the=20 institutional animal welfare committee.=20

In the first set of experiments, ferrets (Mustela=20 furo, female, weight 800-900 g) were infected = intratracheally=20 with either 10³ TCID₅₀or = 10⁴=20 TCID₅₀(n=3D8 for each group) of SARS coronavirus = strain=20 HKU-39849, which had been preincubated in 1 mL of = cell-culture=20 medium at 37=B0C for 1 h with either CR3014 or an human IgG1 = control=20 monoclonal antibody (n=3D4 for each group). The = concentration of=20 antibody in the medium was a 20-fold excess of the = concentration=20 theoretically needed for neutralisation of the challenge = dose (0=B713=20 mg/mL for the lower and 1=B73 mg/mL for the higher challenge = dose).=20 Pharyngeal swabs were taken from the animals before = challenge and on=20 days 2, 4, and 7 after challenge, and frozen at -70=B0C = before being=20 subjected to quantitative RT-PCR. Two animals from each of = the four=20 groups were killed on days 4 and 7 and necropsies were done=20 according to a standard protocol. For assessment of lung=20 inflammation associated with SARS coronavirus infection,=20 haematoxylin-eosin-stained sections from the cranial and = caudal=20 parts of the lung were examined for inflammatory foci by = light=20 microscopy with a 10x objective. For measurement of the SARS = coronavirus titre, homogenates were prepared from = 0=B71-0=B73 g of lung=20 tissue pooled from cranial, medial, and caudal parts of the = lung.=20

All control ferrets had high titres of SARS = coronavirus in=20 their lungs on day 4, and lower titres on day 7. Viral = replication=20 was accompanied by multifocal pulmonary lesions affecting = about=20 5-10% of the surface area of the lung. Histologically = lesions=20 consisted mainly of mild alveolar damage as well as = peribronchial=20 and perivascular lymphocyte infiltration. RT-PCR on = pharyngeal swabs=20 showed that virus was shed in the throat throughout day 7. = Animals=20 exposed to the mixture of virus and CR3014 had almost = undetectable=20 titres of SARS coronavirus in the lung, showed no lung = lesions on=20 day 4 or 7, and did not shed SARS coronavirus in their = throats.=20

In the second set of experiments, two groups of = ferrets (n=3D4=20 in each group) received an intraperitoneal injection of = either=20 CR3014 or the control antibody at a concentration of 10 = mg/kg, 24 h=20 before intratracheal infection with 10⁴ = TCID₅₀=20 SARS coronavirus, strain HKU-39849. Venous blood was drawn = before=20 administration of monoclonal antibody, before challenge, and = on day=20 2 after the challenge. We measured the human IgG1 content of = the=20 serum by ELISA, and the neutralising capacity by a = neutralisation=20 assay using the fixed virus-varying serum-dilution format.=20 Pharyngeal swabs were taken from the animals before the = inoculation=20 and on days 2 and 4 after the challenge for quantitative = RT-PCR. All=20 animals were killed on day 4 and lung samples were taken as=20 described for the first set of experiments.=20

The mean SARS coronavirus titre in lung homogenates = of the=20 animals given CR3014 was 3=B73 logs lower than that of the = control=20 animals (95% CI 2=B76-4=B70 logs, p<0=B7001 [Student's = t test];=20 figure 2A). This difference was accompanied by complete = protection=20 from macroscopic lung pathology (p=3D0=B7013 [Wilcoxon = rank-sum test];=20 figure 2C) and a reduction of microscopic lesions compared = with=20 controls, which all showed multifocal lesions on gross = necropsy=20 (figure 2D). Shedding of SARS coronavirus in the throat was=20 completely abolished in three of the four animals treated = with=20 CR3014 (figure 2B). However, in one animal the level of SARS = coronavirus excretion was similar to that noted in the = control=20 group. The concentration of CR3014 in the serum of this = ferret=20 before challenge was less than 5 =B5g/mL, compared with = 65-84 =B5g/mL in=20 the other three animals, suggesting inappropriate antibody=20 administration. Neutralising serum titres in this animal = were less=20 than half of those in the other animals on day 0 (titre of 5 = against=20 100 TCID₅₀), and were not detectable on day 2 = after=20 infection, compared with a titre of 5-10 against 100=20 TCID₅₀ in the other animals on day 2.=20

=20

Figure 2: SARS coronavirus = replication in=20 lung tissue (A), shedding of SARS coronavirus in = pharyngeal=20 secretions (B), and macroscopic lung pathology (C, D) = in=20 ferrets=20

(A) SARS coronavirus titres in lung homogenates = on day=20 4 after challenge shown as TCID₅₀/mL. = Horizontal=20 bars indicate mean values. Dotted line shows lower = detection=20 limit of assay. (B) SARS coronavirus excretion = measured by=20 RT-PCR in pharyngeal swabs expressed as SARS = coronavirus=20 equivalents (eq), which equal TCID₅₀. One = animal in=20 the CR3014 group, indicated by open circles, shed = virus. (C)=20 All CR3014-treated ferrets were free from macroscopic = lung=20 pathology, whereas all control ferrets showed = multifocal=20 lesions. (D) Left formalin-fixed lung (inflated with = formalin=20 by intrabronchial intubation) of a control ferret, = with=20 multifocal lesions (asterisks). =

SARS coronavirus has been detected in nasopharyngeal=20 aspirates of up to 72% of SARS patients, being associated = with=20 increased mortality, and in the lungs of all autopsied = patients by=20 viral isolation or RT-PCR.⁴ On the basis of our = data and=20 the successful prophylaxis of respiratory syncytial virus = disease=20 with a humanised monoclonal antibody (Palivizumab), we = reason that=20 immunoprophylaxis of SARS coronavirus infection with a human = monoclonal antibody might be an option for the control of=20 SARS.⁵ The prophylactic dose we used for CR3014 = (10=20 mg/kg) is less than the 15 mg/kg dose at which Palivizumab = is given=20 intramuscularly to at-risk infants once a month. If CR3014 = reduced=20 the replication of SARS coronavirus in people to the same = extent as=20 in ferrets, and in view of the serum halflife of up to 20 = days for=20 IgG1 in human beings, one intramuscular administration of = CR3014 at=20 the dose used in this study should protect an adult for the = length=20 of at least one to two SARS coronavirus incubation periods = (median=20 4-6 days). Passive immunisation with CR3014 might, = therefore, be a=20 feasible approach to prevent lung manifestations in people = exposed=20 to SARS coronavirus, and prevent person-to-person spread of = the=20 virus by abolishment of viral shedding in pharyngeal = secretions.=20

Contributors=20

J ter Meulen, A B H Bakker, G J Weverling, J = Goudsmit, and A=20 D M E Osterhaus planned the study, analysed data, and wrote = the=20 report. E N van den Brink, J de Kruif, W Preiser, and W = Spaan=20 generated and characterised the recombinant antibody. H R = Gelderblom=20 did electron microscopy. B E E Martina and B L Haagmans did = animal=20 experiments and virus titrations. T Kuiken did pathological=20 analysis. All authors saw and approved the final version of = the=20 manuscript.=20

Conflict of interest statement=20

The animal studies at the Erasmus University, = Rotterdam, were=20 funded by Crucell Holland BV. The sponsor was involved in = the design=20 of the animal studies and in data analysis, but was not = involved in=20 performing the animal studies or the virological or = pathological=20 analysis of the necropsy samples.=20

Acknowledgments=20

SARS coronavirus strain HKU-39849 was provided by = Malik=20 Peiris, Department of Microbiology, University of Hong Kong. = We=20 thank Geert van Amerongen and Roberto Dias d'Ullois for = technical=20 support in the animal studies, and Frank Pistoor = (ViroClinics),=20 James Simon (CoroNovative), and Jindrich Cinatl for = logistical=20 support of the studies. We thank Rob van Lavieren, Freek = Cox, Mandy=20 Jongeneelen, Gabi Bauer, and Andrea M=E4nnel for excellent = technical=20 assistance.=20

1 Guan Y, Zheng BJ, He YQ, et al. Isolation and=20 characterization of viruses related to the SARS coronavirus = from=20 animals in Southern China.=20 = Science 2003; 302: 276-78. [PubMed]=20

2 Subbarao K, McAuliffe J, Vogel L, et al. Prior = infection=20 and passive transfer of neutralizing antibody prevent = replication of=20 severe acute respiratory syndrome coronavirus in the = respiratory=20 tract of mice. J=20 Virol 2004; 78: 3572-77. [PubMed]=20

3 Martina BE, Haagmans BL, Kuiken T, et al. SARS = virus=20 infection in cats and ferrets. Nature 2003; = 425: 915.=20

4 Tsang OT, Chau TN, Choi KW, et al. = Coronavirus-positive=20 nasopharyngeal aspirate as predictor for severe acute = respiratory=20 syndrome mortality. Emerg Inf=20 Dis 2003; 9: 1381-87. [PubMed]=20

5 Parnes C, Guillermin J, Habersang R, et al. = Palivizumab=20 prophylaxis of respiratory syncytial virus disease in = 2000-2001:=20 results from The Palivizumab Outcomes Registry. = Pediatr=20 = Pulmonol 2003; 35: 484-89. [PubMed]=20

Crucell Holland BV, Leiden, Netherlands (J ter = Meulen MD,=20 A B H Bakker PhD, E N van den Brink PhD, G J Weverling MD, J = de=20 Kruif PhD, J Goudsmit MD); Institute of Medical Virology, = Johann=20 Wolfgang Goethe University, Frankfurt am Main, Germany = (W=20 Preiser MD); Robert Koch-Institut, Berlin, Germany (H = R=20 Gelderblom MD), Department of Medical Microbiology, = Leiden=20 University Medical Centre, Leiden, Netherlands (W Spaan = PhD);=20 Department of Virology, Erasmus Medical Center, Rotterdam,=20 Netherlands (B E E Martina PhD, B L Haagmans PhD, T = Kuiken PhD,=20 A D M E Osterhaus PhD)

Correspondence to: Dr J ter Meulen, Crucell = Holland BV, PO=20 Box 2048, 2301 CA Leiden, Netherlands (e-mail:j.termeulen@crucell.com)<= BR=20 clear=3Dall>