Dynamic Detection of Cytomegalovirus in Breastmilk

Towards a Device for Self Monitoring Risks of Postnatal Infection

S. Py

, A. Guitton

, F. Lardet-Vieudrin

, N. Marthouret

, L. Pazart

, A. Coaquette

, W. Boireau

G. Thiriez

, G. Herbein

3, 5

and B. Wacogne

1, 2

INSERM-CIC 1431, Besançon University Hospital, Besançon, France

FEMTO-ST Institute, University of Bourgogne Franche-Comté, CNRS, Besançon, France

Laboratory of Virology, Besançon University Hospital, Besançon, France

Department of Neonatal Medicine, Besançon University Hospital, Besançon, France

UPRES EA4266, SFR FED 4234, Pathogens and Inflammation Laboratory, Department of Virology,

University of Bourgogne Franche-Comté, Besançon, France

Keywords: Cytomegalovirus, Screening, Biochips, Preterm Infants, Breastfeeding, Self-Testing.

Abstract: Human cytomegalovirus (HCMV) infection is a major cause of morbidity worldwide especially in newborn

infants. While congenital HCMV infection affects 2-5% of preterm newborns, the risk of postnatal infection

particularly through breast milk is higher in this population (prevalence about 20%) since more than one

mother on two is affected. Congenital and postnatal infection can lead to important clinical complications

such as deafness, learning disabilities, and mental retardation during childhood. Neonatologists are squeezed

in their clinical practice: either breastfeeding is favored without any milk treatment going on exposure of

preterm infants to a potential infection, or milk is systematically treated by freezing or pasteurization but

with deprivation of non-at-risk infants from the benefits of fresh milk. In this position paper, we propose a

possible solution to differentiate milk with risk of HCMV contamination from milk without any risk. This

would allow subsequent adaptation of the milk feeding strategy. Also, because the HCMV contamination

peak appears 4 to 8 weeks after birth, the work we present here should lead to a device meant to be used

both at hospital and at home in a self-testing manner.

1 INTRODUCTION

Human cytomegalovirus (HCMV), rarely dangerous

for immune-competent person, is a real threat for

immune-depressed people (organ transplanted or

pregnant women). HCMV is the most frequent

etiologic agent of congenital and postnatal infection

of newborns and can have a significant impact on

the neurosensory development of newborns and

especially preterm infants (Hayashi et al. 2011).

Recent studies showed that postnatal HCMV

infection in preterm infants can lead to serious

clinical consequences and can lead to death in rare

cases (Lanzieri et al. 2013; Hamele et al. 2010;

Hamprecht et al. 2008). Although the long-term

follow-up of the neurosensory development of

congenitally infected preterm infants is well

documented, very few studies concern postnatal

infected preterm infants. However, actual data

suggest a negative influence on long-term cognitive

development (Kurath et al. 2010; Bevot et al. 2012;

Goelz et al. 2013).

Breastfeeding is now clearly recognized as being

superior to artificial feeding. However, breastfeeding

plays a major role in the epidemiology of

transmission and postnatal HCMV infection. It is

now well established that HCMV is excreted in milk

from about 80% of seropositive lactating mothers

(Kurath et al. 2010). Excretion can start since the

first post-partum week and reaches a maximum

value 4 to 8 weeks after birth. Mother-to-child

transmission generally occurs during this period

(Hamprecht et al. 2003, 2008). About 20% of

breastfed children are HCMV positive and the

200

Py, S., Guitton, A., Lardet-Vieudrin, F., Marthouret, N., Pazart, L., Coaquette, A., Boireau, W., Thiriez, G., Herbein, G. and Wacogne, B.

Dynamic Detection of Cytomegalovirus in Breastmilk - Towards a Device for Self Monitoring Risks of Postnatal Infection.

DOI: 10.5220/0006634602000205

In Proceedings of the 11th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2018) - Volume 1: BIODEVICES, pages 200-205

ISBN: 978-989-758-277-6

contamination risk increases with lactating duration.

For preterm infants, the weak transmission of

mother antibodies and the non-mature immune

system increases the risk of symptomatic HCMV

infection.

Today, almost no national recommendations on

the manipulation of the breast milk of HCMV

positive mothers are proposed. Methods exist to treat

breast milk. Systematic pasteurization is not done

because it alters the immune components of milk

(Chang et al. 2013). Freezing at -20 °C does not

completely destroy the virus (Yoo et al. 2015).

Neonatologists are then squeezed in their clinical

practice between the potential risk to transmit

infection when breast milk is not treated and the risk

to favor complications if the milk is treated. The

ideal solution would be to differentiate “at risk” and

“non at risk” situations in order to treat only the “at

risk” cases. This would be extremely more

satisfactory than a systematic attitude.

The techniques currently available to detect

HCMV are Polymerase Chain Reaction (PCR) and

cell culture but are not adapted to rapid and early

HCMV detection in breast milk (Hamprecht et al.

2008). Indeed, PCR is expensive and time

consuming and cell culture gives a result only

several days after sampling. Enzyme-linked

immunosorbent assay (ELISA) can hardly be used

because in this static fluidic configuration, nutritive

components of the milk shield the antibodies used

for detection.

In order to help clinicians in their practice, our

project aims at developing a simple, fast and low-

cost system namely a rapid diagnostic test (RDT)

which could be used at the hospital or at home in a

self-evaluation manner.

The biological and technological hypotheses

developed in this position paper concern the fact that

in a fluid flow configuration, HCMV remains

available for bio-recognition by specific antibodies.

Using specifically marked detection antibodies, an

optical measurement is possible. Prior to the

development of a RDT, a technico-clinical study is

ongoing to develop an integrated immuno-combined

device able to detect HCMV in native breastmilk.

The goal is to determine the most suitable antibodies

association to be used. In section 2, we show that a

fluid flow configuration is much more efficient than

ELISA-like methods although the laboratory model

we used suffered from instabilities. In section 3, we

present the very first results obtained with a stable

and reliable integrated device. In part 4, and in line

with the scope of a position paper, we present

scientific and socio-economic impacts the foreseen

RDT (still to be fabricated) could address.

2 PRINCIPLE AND

PRELIMINARY EXPERIMENTS

2.1 Biochip Principle

The technique we propose to detect HCMV is based

on antigen/antibody recognition. Biochips were

designed and homemade. The biosensor consists of a

polystyrene biochip coated by human polyclonal

anti-HCMV antibodies as shown in figure 1. HCMV

potentially present in the breast milk sample is

captured by these antibodies. HCMV detection uses

a specific secondary antibody coupled to horseradish

peroxidase (HRP) enzyme which subsequently

recognizes the captured virus. After addition of a

substrate of this enzyme, a colorimetric reaction

occurs and allows transforming the substrate to a

blue product. Then, the optical reading relies on an

absorbance measurement at about 640 nm or 450 nm

if the reaction is stopped with sulfuric acid.

Figure 1: Principle of the HCMV biosensor.

2.2 Preliminary Tests

The first HCMV detection tests were conducted with

an initial version of a homemade laboratory model

in parallel with ELISA experiments. The goal was to

compare dynamic (laboratory model) and static

(ELISA) conditions in terms of virus detection. The

laboratory model consists of a fluidic system

containing a biosensor inserted into a cartridge.

Syringes contain reagents which are driven on

biosensor surface. The system allows controlling the

fluid flows and interaction durations. The biosensor

is sandwiched between a light-emitting diode (LED)

emitting in the red wavelength region and a

photodiode as shown in figure 2(a). Initially, the

device was designed for 4 simultaneous tests.

Preliminary positive results were obtained with

the device from artificially contaminated breast milk

samples. Indeed, virus concentrations as low as 6

µg/mL can be detected using this simple opto-fluidic

model with a higher absorbance value than the one

Dynamic Detection of Cytomegalovirus in Breastmilk - Towards a Device for Self Monitoring Risks of Postnatal Infection

201

obtained with ELISA technique (figure 2(b)).

Furthermore, it can be seen that absorbance levels

achieved in ELISA are of the same order of

magnitude as those obtained using irrelevant

antibodies. This highlights the fact that a static

configuration is not suitable for virus detection in

native breastmilk.

Figure 2: HCMV (~6 µg/mL) detection in artificially

contaminated breastmilk in static conditions (ELISA) and

dynamic conditions (laboratory model). Two different

experiments performed with two distinct batches of

capture antibodies are presented.

In view of the first results obtained using

breastmilk, dynamic conditions proved to be

extremely efficient for HCMV detection. Indeed,

optical densities as high as 0.7 are achieved in

dynamic conditions against 0.1 maximum in static

condition. As previously mentioned, the high

proportion of lipids in this biological fluid makes

capture and detection of viruses more difficult due to

a shield effect of the components. These results

should be validated using naturally HCMV

contaminated breastmilk sample and with a much

larger number of samples. Indeed, for this position

paper, only a few experiments have been conducted.

We still don't have quantitative estimation of the

reliability and repeatability.

The laboratory model presented in figure 2

proved to be inappropriate for a large scale study.

Indeed, we experienced number of experimental

problems. The mechanical stability was not high

enough to ensure reproducible measurements. The

gap between the LED and the photodiode is open to

air and variations in the ambient light jeopardize

measurements. Optical densities presented in figure

2(b) were measured after the experiment was

finished by means of an optical spectrometer.

Therefore, no continuous absorption measurements

were possible.

In fact, the first laboratory model was a copy of

the model used to experiment on red cell immuno-

capture (Charrière et al. 2012). Biological reactions

involved in these experiments lasts for a few

minutes only. Optical absorption of red cells is

orders of magnitude higher than the absorption

measured in the current experiments. Also,

experimentation times were a few minutes for red

cells and more than 30 minutes for HCMV.

Therefore, variations of the optical power emitted by

the LED become an issue. Stabilization of the

emitted power is then required in our experiments.

This is why we designed a new, compact and opto-

mechanically stable prototype which allows

continuously recording optical absorption during the

experiment.

3 A STABILIZED AND

INTEGRATED PROTOTYPE

3.1 Description of the New Prototype

A Computer Aided Design of the new prototype is

presented in figure 3. A reaction chamber defined in

a sheet of polydimethylsiloxane is sandwiched

between two biosensors which are used to enhance

the sensor’s sensitivity. Sample and reagents are

introduced in and evacuated from the reaction

chamber through commercially available

microfluidic connectors. Fluid flows are controlled

using motorized syringes. Fluidic sealing is ensured

using two micro O-rings. Embedded spacers are

used to ensure that the thickness of the reaction

chamber (few tens of µm) remains constant between

experiments (new biosensors must be used for each

experiment). Also, embedded stops are used to

ensure a vertical repositioning between experiments.

Stops and spacers are not visible in the figure.

BIODEVICES 2018 - 11th International Conference on Biomedical Electronics and Devices

202

Figure 3: CAD view of the new opto-fluidic prototype.

The optical absorption measurement is provided

by the couple LED plus D1 photodiode. The LED is

embedded in a heat sink in order to reduce self-

heating due to long operations (more than 30 min).

Indeed, LED heating results in a continuous

decrease of its emitted power. Although the heat

sink cannot completely stabilized the LED

temperature, an optically stable operation is ensured

with an embedded electronic optical power

regulator. The beam sensing D2 photodiode is

inserted into a spacer. A 2 mm diameter hole is

drilled in the spacer. It allows collecting only light

which has propagated in straight line from the LED.

Isolation from ambient light is obtained using a large

diameter O-ring placed in a circular groove.

Figure 4: Schematic representation of the driving and

detection electronics.

Driving and detection electronics uses surface

mount component on printed circuit boards (PCBs)

directly integrated in the device. Internal wiring is

not shown in the figure. The D1 photodiode is

integrated in the top part of the prototype while the

optical power regulator and PCBs are integrated in

the bottom part. The general electronic circuit is

schematically described in figure 4. D1 is the

photodiode which produces a current proportional to

the absorption signal while D2 is the photodiode

dedicated to the LED power regulation. Figure 5

shows the actual prototype with a closer view to the

reaction chamber. Stops and light shield are also

visible.

Figure 5: Pictures of the prototype.

3.2 Sample Preparation

For this set of experiments, breastmilk samples were

not available. Experiments were then conducted

using a simplified biological model. Biochips and

samples were prepared as follows. Three biochips

were coated by human polyclonal anti-HCMV

antibodies at a concentration of 20 ng/µL in

carbonate/bicarbonate buffer overnight at 4°C. The

next day, a rinsing of chips with phosphate buffered

saline (PBS) 1X followed by a saturation step of the

surface with Bovine Serum Albumin 10% during 1 h

at room temperature (RT) was performed. A mixture

of commercial HCMV antigen (pure or diluted at

1/25) and an anti-HCMV antibody conjugated to

HRP diluted at 1/5000 (ETI-CYTOK-M reverse plus

Diasorin kit) was injected on two of the chips and

incubated 1 h at RT. A negative control was realized

by incubation of one chip with the antibody alone

diluted in PBS. Here, we call CMVneg the negative

control solution, CMV1/25 the test solution with the

antigen diluted at 1/25 and CMVpure the test

solution with no dilution of the antigen. Three

washing of 1 mL were realised with a wash solution

composed of PBS-Tween (ETI-CYTOK-M reverse

plus, Diasorin) and the HRP substrate (hydrogen

peroxide and tetramethylbenzidine) was incubated

during about 50 min at RT in the prototype.

Absorption measurements were collected during this

time of incubation.

Dynamic Detection of Cytomegalovirus in Breastmilk - Towards a Device for Self Monitoring Risks of Postnatal Infection

203

3.3 Experimental Results Obtained

with the New Prototype

Figure 6 shows the photodiode signal recorded over

the whole experiment for the CMVneg, CMV1/25

and CMVpure samples.

Figure 6: Absorption recorded during the whole

experiment session.

It can be observed that the electronic circuits

perfectly regulate the LED emitted power. The noise

level is extremely small as illustrated in the insert of

figure 6. The noise peak to peak level is of the order

of 8 nA which correspond to a root mean square

value less than 2nA. Also, the exponential decay of

the signal is clearly observed.

Figure 7 shows the signal recorded at the end of

the experiments as a function of the dilution levels:

1/1 CMVpure, 1/25 for CMV1/25 and 1/∞ for

CMVneg.

Figure 7: Final signal as a function of the dilution levels.

The blue curve in the figure is an exponential

fitting of the data. Of course, this is a very rough

approximation because of the very few experimental

data. However, it is expected that the viral charge in

the actual samples will be quite low. This means that

only the left part of the curve should be considered

as representative of real situations. Zooming this

region of the figure and reporting the noise level

allows roughly estimating the detection limit in

terms of dilution level (figure 8).

Figure 8: A tentative estimation of the detection limit.

The value of 2.810

-4

represents a sample where

the antigen dilution is 1/3500. However, this is

nothing but a rough estimation obtained using a

simplified biological model. We are now starting a

clinical trial in order to assess the accuracy and the

potential of the opto-fluidic technique we propose. If

successful, a step ahead will have been passed in the

direction of a home-use RDT. The impacts of such a

device are summarized in the next section.

4 SOCIO-ECONOMICAL

IMPACTS

As mentioned above and in the scope of a position

paper, we present the scientific and socio-economic

impacts such a device potentially.

HCMV is an opportunistic virus that infects a

large proportion of the population worldwide and

results in an asymptomatic latent infection in healthy

subjects. The disease burden is both medical and

economic. HCMV infection can lead to severe

diseases in the absence of an effective immune

response. HCMV is also the leading cause of

neonatal viral infection and can have a significant

impact on the neurosensory development of

newborns and especially preterm infants. HCMV

infection may result from maternal-fetal

transmission during pregnancy (2-5% of very

premature infants) or postnatal transmission (about

20% of children). Currently, viral status of breast

BIODEVICES 2018 - 11th International Conference on Biomedical Electronics and Devices

204

milk is not explored in practice and, depending on

the health centers, milk is systematically inactivated

or breastfeeding is continued with raw milk without

any caution. Finally, although the cost of HCMV

infection in the hospital community has not yet been

clearly established, it appears that HCMV infections

cost hundreds of thousands of euros each year to the

French health system in terms of medical and

surgical expenses, especially in taking care of long-

termed disabled children and adults infected early in

life or during pregnancy.

Therefore, an easy-to-use secured RDT to detect

HCMV infection in breastmilk from lactating

women of preterm infants is urgently needed. An

answer to this need is the subject of this

communication. About 8000 very preterm infants

could benefit from this test each year in France and

13 million worldwide. Considering that the test

should be repeated several times for a same couple

mother/baby pair in the early months of

breastfeeding, the market worth to be taken into

account. Indeed, the test will be practiced both at

hospital and at home since the peak of viral

excretion in breast milk occurs generally after

hospitalization of the child. In addition to detection

by caregivers in departments of neonatal medicine,

self-diagnosis of mothers will be possible given ease

of use and reading of this type of test.

5 CONCLUSIONS

Although the risk of HCMV congenital infection is

relatively low, the risk of postnatal contamination, in

particular via breast milk, can be dramatic for

preterm infants. Currently, the question is: should

we favor a better development and take the risk of

using contaminated breast milk, or should we use

treated milk, even when the HCMV infection is low

enough to be considered safe?

To address this problem, and in the current

context of breastfeeding promotion, we propose to

develop a HCMV biosensor based on sandwich

ELISA principle in a dynamic flow configuration

(lateral flow immunochromatography). This position

paper presents studies that have just started, but we

think it is possible to set-up an easy to use and rapid

"point-of-care" device to detect HCMV in

breastmilk. Therefore, a third answer can be

proposed to the above mentioned question. The idea

is to screen HCMV on a routine basis and to define a

personalized feeding strategy for “at risk”

population only. Without such a rapid HCMV test,

this third solution may never exist.

ACKNOWLEDGMENTS

This work is funded by the “APICHU-RBFC call” in

2015.

The authors would like to thank FEMTO-

engineering for the design and manufacture of the

prototype.

REFERENCES

Bevot, A, et al., 2012. Long-Term Outcome in Preterm

Children with Human Cytomegalovirus Infection

Transmitted via Breast Milk. Acta Paediatrica 101

(4): e167-172.

Chang, J.C., et al., 2013. Influence of Prolonged Storage

Process, Pasteurization, and Heat Treatment on

Biologically-Active Human Milk Proteins. Pediatrics

and Neonatology 54 (6): 360‑66.

Charrière, K., et al., 2012. SmarTTransfuser: A biochip

system for the final ABO compatibility test.

BIODEVICES 2012 - Proceedings of the International

Conference on Biomedical Electronics and Devices,

janvier, 257‑62. SCITEPRESS.

Goelz, R., et al., 2013. Long-Term Cognitive and

Neurological Outcome of Preterm Infants with

Postnatally Acquired CMV Infection through Breast

Milk. Archives of Disease in Childhood. Fetal and

Neonatal Edition 98 (5): F430-433.

Hamele, M., et al., 2010. Severe Morbidity and Mortality

with Breast Milk Associated Cytomegalovirus

Infection. The Pediatric Infectious Disease Journal 29

(1): 84‑86.

Hamprecht, K., et al., 2008. Cytomegalovirus transmission

to preterm infants during lactation. Journal of Clinical

Virology 41 (3): 198‑205.

Hamprecht, K., et al., 2003. Rapid detection and

quantification of cell free cytomegalovirus by a high-

speed centrifugation-based microculture assay:

comparison to longitudinally analyzed viral DNA load

and pp67 late transcript during lactation. Journal of

Clinical Virology 28 (3): 303‑16.

Hayashi, S., et al., 2011. Transmission of cytomegalovirus

via breast milk in extremely premature infants. J

Perinatol 31 (6): 440‑45.

Kurath, S., et al., 2010. Transmission of Cytomegalovirus

via Breast Milk to the Prematurely Born Infant: A

Systematic Review. Clinical Microbiology and

Infection 16 (8): 1172‑78.

Lanzieri, T., et al., 2013. Breast Milk-Acquired

Cytomegalovirus Infection and Disease in VLBW and

Premature Infants. Pediatrics 131 (6): e1937-1945.

Yoo, H.S., et al., 2015. Prevention of Cytomegalovirus

Transmission via Breast Milk in Extremely Low Birth

Weight Infants. Yonsei Medical Journal 56 (4):

998‑1006.

Dynamic Detection of Cytomegalovirus in Breastmilk - Towards a Device for Self Monitoring Risks of Postnatal Infection

205