Technical Field The present invention generally relates to bacterial preparations obtained from blood cultures (hemocultures) and to purification of bacteria present in blood so as to allow the characterisation of the bacteria. More specifically, the present invention relates to a method of treating blood samples, to a method of preparing a bacterial preparation suitable for use in bacterial characterisation, diagnosis and/or taxonomic identification, to methods of diagnosis and/or other types of bacterial characterization, including antibiotic susceptibility testing.

Prior Art and the Problem Underlying the Invention

In many hospitals, outpatient, stationary or emergency patients exhibiting certain symptoms, such as fever, or other symptoms indicating the possibility of a systemic inflammatory response syndrome or sepsis, are routinely subjected to testing for bacterial bloodstream infection.

Generally, blood samples are taken from patients and subjected to hemoculture. If in the course of the liquid hemoculture, bacterial growth is detected, the content of the hemoculture vial is centrifugated and the pellet containing the bacteria, besides further undesired components such as erythrocytes and wastes from the hemoculture process, is placed on an agar plate in a Petri dish and thus subjected to solid culturing. The next day, a sufficient amount of bacteria is obtained and subjected to further tests allowing diagnosis and the choice of the appropriate treatment of the patient. These further tests involve, for example, taxonomic characterisation using, for example, mass spectroscopy, biochemical and antimicrobial evaluations of the bacteria. Following these final tests, the bacteria is generally completely characterised and the treatment of the patient can be adapted according to the results of the bacterial identification and susceptibility testing.

In the above-depicted process taking place between the taking of a blood sample and the adequate medical treatment of a patient, many steps are conducted in automated, computer assisted processes. In the beginning, the blood sample is added to a liquid medium in a special vial or container and placed in an incubator equipped with measurement devices capable of detecting the presence of bacterial growth, generally within one to five days.

Such hemoculture instrumentation, including an incubating system, special vials for aerobic and/or anaerobic culturing, computers, automated microbial detection systems, operator interfaces and the like are commercialised by BioMerieux, France (for example, the BacT/ALERT© device) and by Becton Dickinson, US (for example the BacTec™ 9240 device). The hemoculture process following blood withdrawal generally takes from about two to five days.

Bacterial characterisation, on the other hand, today is generally also accomplished in an automated process. If a sufficient amount of cultivated bacteria is available, the genus and species of the bacterium can be determined, for example by using mass spectroscopy in a MS MALDI-TOF apparatus and comparison of the resulting spectrum with suitable database spectres and/or using commercial biochemical/enzymatic tests. Such biochemical and enzymatic evaluation is today generally conducted in automated, computer assisted processes and devices, such as the Vitek® instrument of BioMerieux, France.

Also the testing of the effect of various antibiotics on the growth of the pathogen (antimicrobial evaluation) may be done using commercial automated computer-assisted processes and devices, such as the Vitek® instrument.

The present invention addresses, in a main aspect, the problem of time spent between the blood withdrawal and availability of the bacterial characterisation. In order to conduct the latter and last step in this process, a sufficient amount of viable bacteria that is substantially free of disturbing elements such as non-bacterial proteins and the like, needs to be provided.

However, the successive steps of liquid hemoculture and solid culture presently employed to this end are time consuming, taking generally from four to seven days. More specifically, one day is generally used to obtain enough bacteria to conduct downstream identification tests. This is particularly detrimental, since the patient from which a contaminated blood sample was taken generally is in a critical state and needs specific treatment as soon as possible. WO 2009/0565580 discloses a method of the mass spectrometric identification of pathogens in body fluids, such as urines and blood. The procedure seems indeed to be adapted to clear body fluids, but only when a large number of bacteria are present. Moreover, the described approach used to remove blood cells, i.e. osmotic destruction with water, is not very efficient and in general the procedure disclosed in this document does not allow for efficient removal of red blood cells. Furthermore, this WO 2009/0565580 document is limited to taxonomic characterisation by mass spectrometry and is not directed to the preparation of viable bacteria for use in antimicrobial susceptibility testing or in biochemical identification (commercial identification gallery or individual enzymatic or phenotypic tests).

Summary of Invention

Remarkably, the present inventors provide a method of preparing bacterial preparations from cultures vials inoculated with blood samples taken from infected patients that is substantially shorter than the methods used previously and thus achieves an important saving of time in the period between blood withdrawal and bacterial characterisation, such as species identification and antibiotic susceptibility testing.

The inventors provide a surprising process, which allows the use of bacteria contained in positive hemoculture suspensions for bacterial characterisation. This was previously impossible, mainly due to the presence of blood and blood components in the hemoculture suspension, but also for other reasons, which will be set out in further detail below.

Accordingly, the present invention provides, in a first aspect, the use of bacteria cultivated solely in a liquid hemoculture in methods of bacterial characterisation and/or diagnosis.

In a second aspect, the present invention provides a method for removing red and/or white blood cells from a hemoculture suspension, said method comprising one or more steps of the methods of the invention. This method may comprise, for example, one or more centrifugation steps and/or the step of adding a treatment solution, all preferably as defined herein.

In a third aspect, the present invention provides a method for producing a bacterial preparation from a blood sample in a form that is suitable for use in bacterial characterisation, diagnosis, antibiotic susceptibility testing and/or taxonomic identification, the method comprising the steps of: adding a blood sample to a liquid medium thereby obtaining a liquid hemoculture starting suspension (the starting suspension); cultivating the starting suspension for a period of 2 hours to 12 weeks under conditions that are conducive to bacterial growth, thereby obtaining a positive cultivated bacterial hemoculture suspension (the hemoculture suspension) if said blood is contaminated; subjecting said hemoculture suspension to a first centrifugation, preferably conducted at a force of 500 * g (g = 9.81 m/s ) or higher, thereby obtaining at least a first supernatant liquid and a first pellet; removing the first supernatant liquid; suspending the first pellet in a treatment solution preferably comprising ammonium ions in order to lyse red blood cells; subjecting the suspended first pellet to a second centrifugation, said second centrifugation taking place at a lower centrifugal force than said first centrifugation, preferably a force of 400* g or lower, thereby obtaining at least a second pellet comprising viable bacteria and a second supernatant liquid; removing the second supernatant liquid; and optionally, conducting one or more washing and/or extraction steps with said second pellet.

In a fourth aspect, the present invention provides a method of identifying and/or characterizing bacterial species present in a blood sample, the method comprising the steps of: providing a liquid positive cultured bacterial hemoculture suspension (the hemoculture suspension) cultivated for 2 hours to 12 weeks under conditions conducive to bacterial growth, said hemoculture suspension being found positive for bacterial contamination; subjecting the hemoculture suspension to a first centrifugation step, thereby obtaining at least a first pellet and a first liquid supernatant; removing the first liquid supernatant; suspending the first pellet in a lysis solution conducive to the lysis of blood cells, in particular erythrocytes contained in the pellet; subjecting the suspended first pellet to a second centrifugation, thereby obtaining at least a second pellet and a second liquid supernatant; removing the second liquid supernatant; optionally, conducting one or more washing and/or extraction steps with said second pellet; and identifying the bacterial species present in said second pellet by way of mass spectrometry and/or, if an extraction step is absent, by biochemical evaluation.

In a fifth aspect, the present invention provides a method of treating bacteria and/or preparing a bacterial preparation, the method comprising the steps of:

providing a liquid positive cultured bacterial hemoculture suspension; subjecting the hemoculture suspension to at least one step of centrifugation;

exposing the hemoculture suspension and/or the centrifugated hemoculture suspension to a treatment solution; and,

separating bacteria contained in the treatment solution from liquid and/or other undesired components.

In a sixth aspect, the present invention provides a method for diagnosing a bacterial infection of the blood of a patient, the method comprising the steps of:

conducting a hemoculture comprising a blood sample of a patient and a liquid medium thereby obtaining a hemoculture suspension;

subjecting the hemoculture suspension to at least one step of centrifugation;

exposing the hemoculture suspension and/or the centrifugated hemoculture suspension to a treatment solution; and,

separating bacteria contained in the treatment solution from liquid and/or other undesired components thereby obtaining an bacterial preparation suitable for automated analysis; and,

subjecting the bacterial preparation to automated characterization thereby diagnosing the bacterial infection. In a seventh aspect, the present invention provides a method of producing a bacterial preparation from a positive cultured hemoculture suspension, said method comprising the steps of the method according to the first, second, third, and/or fourth aspect of the invention.

In an eighth aspect, the present invention provides a kit, said kit comprising a treatment solution as defined herein, or a concentrate thereof, and possibly a protocol of the steps as defined herein.

In a ninth aspect, the present invention provides the use of a treatment solution for removing red and/or white blood cells from a hemoculture suspension.

Further aspects and preferred embodiments of the aspects will be set out herein below. Brief Description of the Drawings

Figure 1 shows flasks with positive blood cultures with Staphylococcus aureus (aerobic vial). At the bottom of the flasks, a hemorrhagic pellet obtained after initial centrifugation can be seen.

Figure 2 shows the same cultures as Figure 1, with the supernatant being removed and the pellet resuspended in a treatment solution of the invention (left tube) or in the isotonic NaCl negative control (right tube). The figures show pellets obtained following a second centrifugation.

Figure 3 shows bacterial pellets obtained following removal of the treatment solution of the invention or the isotonic NaCl control solution, as described in Figure 2. The pellet on the left, obtained in accordance with the present invention, can be used directly for MALDI-TOF analysis or any other downstream identification/characterization assays, including antibiotic susceptibility testing. The right tube contains a large number of erythrocytes in the pellet and cannot be used for these analyses.

Detailed Description of the Preferred Embodiments

The present invention relates to a method of treating blood samples, to a method or preparing a bacterial preparation and to further methods, uses and also to a kit. Reference to a method in the present specification is a reference to all methods and also to the uses and kits, and vice versa, of the invention.

The present invention also provides a method of identifying and/or characterizing bacterial species initially present in a blood sample of an individual and amplified by culture in a blood culture broth. In some embodiments of the invention, the step of amplifying the bacteria by hemoculture is also encompassed.

The present invention generally relates to the use of a positive bacterial hemoculture suspension (the hemoculture suspension) in the preparation of a bacterial preparation that can be further used for characterisation of the bacteria. Steps directed to the preparation and/or conducting the hemoculture may be part of certain aspects of the invention or may not be part. Similarly, steps directed the characterisation of the bacterial preparation may also be part of certain aspects of the methods of the invention. The bacterial preparation obtained at the end of the methods of some aspects of the invention is preferably a viable preparation, unless protein extraction steps are conducted as part of the methods of the invention.

According to an embodiment, the method of the present invention comprises the step of adding a blood sample to a liquid medium thereby obtaining a liquid hemoculture starting suspension, also referred to herein as the "starting suspension". The liquid medium may in particular be a blood culture broth. When the sample was taken from an adult, generally about 3 to 12, preferably 5-10, more preferably 6-9 ml blood is added to a container or bottle comprising about 30-60, preferably 40-55 ml liquid medium. When the sample was taken from a young child or even an infant, much less blood can be taken (about 1 to 3 ml is generally recommended). Roughly, the mixing ratio of blood to liquid medium is about 1: 10. The terms "comprise" and "comprising" and the different verbal forms thereof are intended to mean "include, amongst others". They are not understood to mean "consists only of".

The generally bottle-like containers to which the blood sample is added are generally provided by the provider of hemoculture or incubator systems, such as BioMerieux, France and Becton Dickinson, US, as mentioned already above. These containers generally have special characteristics, which allow the monitoring of the contents of the container for biological activity. Accordingly, the containers are generally equipped with a semipermeable membrane, generally at the bottom of the container, which allows a detection system to monitor pH, redox potential and/or C0 ₂ development, for example, inside the container. The container is generally placed manually into the incubator and put under automated monitoring by the data processing unit of the system. Inside the incubator, the container is subjected to shaking, so as to provide optimal growth conditions. Generally, different containers are used for aerobic, anaerobic cultures and cultures conducted with blood taken from infants. The containers generally contain specific labels, for example with clear colour codes, which allow distinguishing between the different kinds of hemoculture.

The containers, which may be mainly made from glass, for example, generally contain an antibiotic absorption component. The antibiotic absorption component has the purpose of absorbing antibiotics which putatively are present in the blood sample, for example if the patient from whom the sample was taken was already under antibiotic treatment. The different providers of hemoculture systems (BioMerieux, Becton Dickinson) generally use different antibiotic absorption components. Containers of some providers contain activated carbon, also referred to as activated coal. Other providers use resin-based beads for absorbing antibiotics.

It is now surprising to note that the method of the present invention works less well or does even not work at all if activated charcoal (activated carbon) is used in the hemoculture container. The inventors do not have a specific explanation why this is the case. Without wishing to be bound by theory, the active carbon is generally used to absorb antibiotics and it is hypothesized that these charcoal particles exhibit a wide size distribution and consequently, all particles cannot be readily separated by centrifugation as specified herein.

Therefore, according to an embodiment of the present invention, said step of cultivating the starting suspension under conditions that are conducive to bacterial growth is conducted in a container that is free of activated carbon, and/or wherein said hemoculture suspension is cultivated in a container that is free of activated carbon and/or activated coal. Preferably, the container comprises a resin for absorbing undesired compounds such as antibiotics.

According to an embodiment, the method of the present invention comprises the step of cultivating the starting suspension thereby conducting said hemoculture. Preferably, the step of cultivating is performed at conditions that are conducive to bacterial growth. Such conditions are known to the skilled person and include, besides a temperature of preferably about 34.5 to 38.5°C and regular shaking, the use of liquid media which contain nutrients that can be used by all bacteria. Samples of each patient are generally separately grown under aerobic and anaerobic conditions.

The step of cultivating the starting suspension is preferably conducted for at least 1, preferably at least 4, more preferably for at least 10 hours, for example for 2 hours to 12 weeks, preferably up to 6 weeks. According to an embodiment, the step of cultivating is conducted for 12 hours to 6 weeks, for example 12 to 125 hours, preferably 24 to 125 hours, most preferably from 36 to 100 hours. Generally, the step of cultivating is conducted until the hemoculture system automatically detects the presence of bacterial activity. In the present specification, this is also referred to by the phrase "said hemoculture being found positive". Generally, it is the automated, computer-assisted hemoculture system, which "finds" or detects the presence of contamination. Theoretically, it contamination could also be found or confirmed by visual inspection, as was done formerly.

It is noted that the time intervals mentioned above span long periods of time, which is due to the large differences in growing rates of different bacterial organisms in hemocultures, some of which grow very slowly.

The terms "contaminated" and or "contamination" and the various grammatical forms thereof, as used in the present specification, refer to the presence of bacterial colony forming units (CFU) in the blood of the patient from which the blood sample was taken. Such presence generally is considered to be a sign of bacterial infection.

If the starting suspension was contaminated (that is, the blood of the patient contains bacteria), a positive cultivated bacterial hemoculture suspension is obtained. This positive hemoculture is also referred to herein as the "hemoculture suspension". The hemoculture suspension is thus a bacterial culture obtained from cultivation of a medium comprising blood and/or blood components.

Generally, if the blood sample contained bacteria, this is detected within about 2 to 4 days following the start of the cultivation step. If after five days no bacterial growth is detected, the hemoculture is generally stopped and the container with the negative blood sample is discarded.

For the purpose of the present specification, the term "cultivating" or "culture" refers to specific steps or methods under conditions that are conducive to bacterial growth and in the presence of a culture medium that is specifically added to promote bacterial growth. Potential growth of bacteria taking place in the course of other steps described herein, or between such steps, for example before adding a blood sample to a hemoculture container or during or following a centrifugation or other treatment step is not considered herein as "cultivating" or "culture".

The (positive) hemoculture suspension obtained after the step of cultivating the starting suspension preferably comprises at least 10 2 , more preferably 5*102 , even more preferably 103 , preferably at least 2*103 , even more preferably at least 3*103 and most preferably at least 5*10 CFU (colony forming units).

According to an embodiment, the method of the present invention comprises the step of subjecting said hemoculture suspension to a first centrifugation, thereby obtaining at least a first supernatant liquid and a first pellet.

According to an embodiment, the method of the present invention comprises, before subjecting the hemoculture suspension to a first centrifugation, adding water or another suspension with low protein concentration (< 5g/l) to the hemoculture suspension. For example, a suitable solution, such as an adequate buffer solution can be used. Water, preferably sterile water, can most conveniently be used for this step, which does not exclude the use of other buffers, solutions and suspensions for this step. Preferably, to about 3-10 ml, preferably 4-7 ml of the hemoculture suspension 25-70 ml, preferably 30-50 ml water is added. Preferably, sterile H ₂ 0 is added.

Preferably, said first centrifugation is conducted at a first speed and/or at a first centrifugal force. According to an embodiment, said first centrifugation is conducted at a force of 400 * g (also expressed herein as "400g"), wherein "*" refers to multiplication ("times") and g is the earth gravitational acceleration (9.81 m/s ). Preferably, said first centrifugation is conducted at a force of at least 500g, more preferably at least 600g, 700g, 800g, 900g and most preferably at least lOOOg. The first centrifugation is preferably conducted for 4-30 minutes, more preferably 7-20 minutes and most preferably 9-15 minutes. Preferably, said first centrifugation is conducted at a centrifugal force that is not higher than 10'OOOg, more preferably not higher than 6'000g, most preferably not higher than 3'000g.

The first centrifugation may thus be conducted at centrifugal forces of 400-3000g, for example 700- Γ500.

According to an embodiment, said first centrifugation is conducted at a centrifugal force that is higher than the centrifugal force at which a second centrifugation is conducted, the latter being described in further detail below. The first centrifugation generally results in the formation of a first pellet, basically containing bacteria and blood components, such as erythrocytes, and a first supernatant, basically containing water stemming from the optionally added water, from the serum and from the liquid medium. The supernatant may also comprise water soluble components that were present in the aforementioned liquids.

According to the present invention, the first supernatant is preferably removed, for example by simple decantation. The first supernatant may be discarded. According to an embodiment, the first pellet is suspended in a treatment solution. The treatment solution is preferably conducive to the separation of blood cells, in particular erythrocytes, from bacteria contained in the pellet. The treatment solution preferably has the capacity to lyse red blood cells and/or to remove blood cells, in particular red blood cells from the bacteria. Since the bacterial preparation to be prepared in the method of the invention contains preferably viable bacteria, it is preferred that the supernatant does not destroy the integrity and/or reduce the viability of the bacteria. Therefore, according to an embodiment, the treatment solution is a lysis solution comprising water and components that are suitable to lyse blood cells while preferably leaving the bacteria contained in the pellet viable. Following this step, it is preferably possible to recover viable bacteria.

According to an embodiment, the treatment solution comprises a cationic compound comprising a positively charged nitrogen atom. Preferably, said cationic compound comprising a positively charged nitrogen atom is selected from ammonium and substituted ammonium ions, such as primary ammonium compounds, secondary ammonium compounds, tertiary ammonium compounds and quaternary ammonium compounds. According to a preferred embodiment, said cationic compound is selected from ammonium, primary, secondary, and tertiary ammonium compounds, the charge of said compound thus being dependent of the pH of the solution. The cationic compound may thus be selected from organic and inorganic compounds.

The treatment solution may or may not comprise surface active compounds.

According to a preferred embodiment, the treatment solution comprises ammonium ions. The treatment solution is preferably prepared by addition of a salt of the cationic compound mentioned above and an anionic compound. The anionic compound may be an organic or inorganic anion. Preferably, it is an inorganic anion. More preferably, it is a halide, such as chloride, fluoride, bromide, and so forth.

For example, the treatment solution is a solution comprising ammonium chloride.

Preferably, the treatment solution is a 0.05 to 1 Molar solution of NH ₄ C1, preferably a 0.08- 0.5 M, most preferably a 0.1-0.2 M solution of NH ₄ C1, or of another ammonium or aminium salt.

Preferably, the treatment solution comprises potassium ions (K ⁺ ) and/or hydrogen carbonate (HCO ₃ ). Preferably, the treatment solution is, in addition to potential other components, a 0.5-5 mM, preferably a 0.8-4 mM solution of KHCO ₃ . According to an embodiment, the treatment solution solution comprises or substantially consists of ammonium chloride and/or potassium hydrogencarbonate dissolved in water.

Surfactants (surface tensioactive agents) such as polyethylene glycol p-(l,l,3,3- tetramethylbutyl) -phenyl ether (Triton X-100), if present in the treatment solution, could affect the viability of the bacteria so that biochemical evaluation is no longer possible. Therefore, according to an embodiment, the treatment or lysis solution is substantially free of such surfactants in general. In particular, the treatment solution is preferably free of surface tension-active agents other than dissolved ammonium salts or other ammonium compounds as specified elsewhere in this specification. According to an embodiment, the treatment solution is substantially free of non-ionic and/or other ionic surfactants. Preferably, the treatment solution is substantially free of polyethylene glycol p-(l,l,3,3-tetramethylbutyl)-phenyl ether (Triton X-100) and of closely related compounds. According to an embodiment, the treatment solution is free of saponins and/or SDS (sodium lauryl sulfate). On the other hand, while surfactants may affect the viability of bacteria, this need not necessarily prevent bacterial identification by mass spectrometry in accordance with the invention. Therefore, according to another embodiment, the treatment solution comprises surfactants, in particular other or additional surfactants than dissolved ammonium salts or other ammonium compounds as specified elsewhere in this specification. According to this particular embodiment, the treatment solution comprises any one or a mixture of several selected from polyethylene glycol p-(l,l,3,3-tetramethylbutyl)-phenyl ether (Triton X-100), saponins and/or SDS (sodium lauryl sulfate), besides, possibly, a dissolved ammonium salt. For example, the pellet obtained following removal of the first supernatant is suspended in 0.3-5, preferably 0.7-2 ml of the treatment solution.

According to an embodiment, the present invention comprises the step of subjecting the suspended first pellet to a second centrifugation. The second centrifugation generally results in the separation of a second pellet comprising bacteria, preferably devoid of blood components, and a second supernatant, generally containing the treatment solution and blood components, in particular erythrocytes and debris thereof.

Preferably, the second centrifugation is conducted at a second speed and/or a second centrifugal force. According to an embodiment, the second centrifugation is conducted at a lower centrifugal force than said first centrifugation. Preferably, said second centrifugation is conducted at a force of 400g or lower, 300g or 200g or lower. Most preferably, the second centrifugation is conducted at a centrifugal force of 50-300g, preferably 80-250, more preferably 100-200g, most preferably 120-160g.

According to an embodiment, the method of the present invention comprises the step of removing the second supernatant, which is generally in the form of a liquid. The second supernatant may be removed by decantation. It may be discarded. The second pellet may be used for bacterial characterisation as detailed further below.

If, however, the second pellet is still found to contain traces of blood (if it is "hemorrhagic"), one or several further washing steps may be conducted. Washing steps may be conducted by adding water or a treatment solution to the second pellet. For example 0.5-7 ml water may be added to the second pellet and the second pellet is suspended therein. If treatment solution is used, the treatment solution as defined above or preferably in a less concentrated form may be used. Thereafter, the suspended pellet is again subjected to centrifugation, similar to the second centrifugation depicted above.

If the viability of the bacteria at this stage is no longer necessary, for example for the purpose of taxonomic characterisation by way of mass spectroscopy, proteins may be extracted from the bacterial pellet. Proteins may be extracted as is conventional, for example mixing the bacterial pellet (the second pellet, optionally further washed) with ethanol, further centrifugation at, for example at about 8'000-15'000g for 0.5-10 minutes, mixing the pellet with formic acid and acetonitrile, centrifugating again at about 8'000-15'000g for 0.5-10 minutes, and transferring a small sample of protein extract onto a target plate and letting it dry.

However, in accordance with the present invention it is also possible to use the (optionally washed) pellet following the second centrifugation and supernatant removal directly for analysis by mass spectroscopy, in particular MALDI- TOF MS, without an extraction step such as described above.

Generally, a MALDI matrix is applied to the extracted and/or unextracted bacterial samples. An exemplary MALDI matrix comprises a-cyano-4-hydroxy-cinnamic acid in acetonitrile and trifluoroacetic acid.

According to an embodiment, the taxonomic genus, more preferably the species of the bacterium is determined, preferably by mass spectroscopy, including the comparison of the spectrum obtained with spectres of a database. An exemplary MALDI-TOF MS device, analysis software, comparison database and identification threshold levels are provided in the example further below.

According to an embodiment, the method further comprises the step of identification of the bacterial species contained in the second pellet by mass spectrometric analysis, optionally following a step of extracting proteins from said second pellet.

The (optionally washed) second pellet, or a part thereof, may also be used in other or further bacterial characterization methods and/or diagnosis. It is noted in this regard that the bacteria contained in the second pellet are alive (viable) and thus suitable for use in biochemical and/or antibiotic evaluation.

Standard biochemical identification is conducted in tests of oxidase activity, catalase activity and commercial identification strips, for example.

Automated characterization in devices commercialized by BioMerieux, France (for example Vitek® and API®), includes growing tests in the presence of different carbohydrate substrates and/or antimicrobial tests. In both cases, a bacterial sample is divided in a plurality of subsamples, all grown (a) in the presence of different carbohydrate substrates and/or (b) in the presence of a different antibiotic agent. This process is conducted in an automated, computer assisted manner, and produces as a direct result on an output medium (monitor, out print and the like) in the case of (a) a list of all carbohydrates on which the bacterium can grow and an according taxonomic determination or suggestion, or in the case of (b), a list of antibiotics to which the respective bacterium is sensitive.

According to an embodiment, the method of the invention comprises the step of identification of the bacterial species contained in the second pellet by biochemical evaluation. According to an embodiment, the method of the invention comprises the step of subjecting the living bacteria contained in the second pellet to automated antibiotic susceptibility testing.

The above described characterization, in particular as described under (b) above, allows the most efficient and effective treatment of the patient from which the contaminated blood sample was taken, by using the most appropriate antibiotic.

It is noted that, according to an embodiment, the present invention is preferably not intended to prepare bacterial preparations and/or pellets of mycobacteria. Mycobacteria have particular cell wall components and are generally cultivated in separate and particularly adapted blood cultures. The extractions of proteins in mycobacteria are also particular. Currently, it is not envisaged to provide a simple, unified procedure for producing a bacterial preparation from a blood sample in a form that is suitable for all bacterial taxa together, including mycobacteria, for use in bacterial characterization, diagnosis and/or taxonomic identification. In general, mycobacteria require particular treatments. The methods of the present invention differ from the methods of the prior art previously used in several aspects. The important advantage of time saving is achieved basically because the method of the present invention does not require culturing bacteria obtained from hemoculture on a solid medium. Such a step is thus preferably absent in the methods of the present invention. Cultivation of bacterial preparations obtained from hemoculture according to prior art methods on solid media generally takes several hours to a few days. Only after this comparatively long time span a sufficiently large number of bacteria sufficiently pure to be useful for bacterial characterisation and/or diagnosis is available. Further steps as the first and/or second centrifugation as disclosed herein and or the treatment with a treatment solution are not used or known in corresponding methods of the prior art.

The present invention may be provided in the form of a kit. The kit may comprise a treatment solution as defined herein, or a concentrate thereof. The kit may further comprise instructions or a protocol for conducting one or more steps, for example centrifugation steps, or the complete methods as described herein.

The invention is illustrated by the examples below, which are not intended to limit the scope of the invention. Examples

Blood cultures are the best approach to establish the etiology of bloodstream infections and infectious endocarditis. Moreover, rapid identification of etiological agent of such severe infections are pivotal to guide antimicrobial therapy. Thus, the impact of timely microbiology laboratory reporting is maximal at the notification of positive blood cultures. The matrix- assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) allows the identification at the species level in few minutes of both Gram positive (1, 2) and Gram negative bacteria (3, 4) by measuring molecular masses of proteins from whole bacterial extracts (references are listed below).

Example 1 : Preparation of a bacterial pellet and protein extraction

We applied a novel, rapid procedure for lysing erythrocytes from positive blood cultures and prepared a bacterial pellet for MALDI-TOF MS analysis. Pellets from positive blood culture vials (Plus aerobic/F, Lytic anaerobic/F and Peds/F, Becton Dickinson, USA) detected by the automated blood culture system Bactec 9240 (Becton Dickinson) were prepared as follows. Five ml of positive medium was added to 40 ml of sterile H ₂ 0. The sample was mixed and centrifuged at l'OOOxg for 10 min. H ₂ 0 and blood cells were removed (Figure 1). The pellet was then suspended in 1 ml of ammonium chloride lysing home-made solution (0.15 M NH ₄ C1, 1 mM KHC0 ₃ ) and centrifuged at 140xg for 10 min (Figures 2, 3). When the pellet remained hemorrhagic, the supernatant was discarded and the pellet was washed again with 2 ml of H ₂ 0. MALDI-TOF MS analysis was then directly performed on the bacterial pellet or after an additional extraction step.

To extract proteins, 20 μΐ of the pellet was mixed with 1 ml of ethanol 70%. After a further centrifugation at 13'000xg for 2 min, the pellet was mixed with 25 μΐ of 70% formic acid and 25 μΐ of pure acetonitrile. After centrifugation at 13'000xg for 2 min, 1 μΐ of the supernatant containing the bacterial extract was transferred onto the MALDI target plate and dried. Subsequently, both unextracted and extracted samples were overlaid with 1 μΐ of MALDI matrix (a saturated solution of alpha-cyano-4-hydroxy-cinnamic acid in 50% acetonitrile- 2.5% trifluoroacetic acid) and dried in air.

Example 2: Species identification using mass spectrometry and verification

Mass spectra were then acquired using the Microflex MALDI-TOF MS (Bruker Daltonics, Bremen, Germany). MALDI BioTyper 2.0 software was used for spectra analysis and comparison with the MALDI BioTyper database. The identification was considered as valid at the species level when the score value was >2, as valid at the genus level when the score value was >1.7 and < 2 and as not valid when the score was < 1.7. The identifications obtained with MALDI-TOF MS analysis were compared with standard biochemical identification with oxidase, catalase, and commercial identification strips, i.e. Vitek (bioMerieux, France) and API (bioMerieux). During the study period, 126 positive blood vials from 78 patients were analyzed. Four cases were excluded from the analysis since bacteremia were polymicrobial. Table 1 shows the results of MALDI-TOF MS identification for the 122 remaining analyses. Among these 122 positive blood cultures, as many as 96 (78.7%) bacterial identifications were obtained by MALDI-TOF MS analysis, of which 69 (56.6% of 122) exhibited a score value >2 and 27 (22.1%) a score value >1.7 and < 2. Importantly, 95 (98.95%) of the 96 bacterial identification were correct at the species level and 1 identification was correct at the genus level only (Staphylococcus pasteuri instead of Staphylococcus caprae). For the latter case, the score value was of 1.73. Thus, among the 27 cases with accurate identification at the genus level, 26 of 27 % (96.3%) were also accurate at the species level. Moreover, among the 69 valid identifications at species level (score above 2), no MALDI-TOF MS results were discordant with conventional identification.

In 26 (21.3%) of cases, no reliable identification were obtained (score value <1.7). As many as 21 of these 26 bacteria (80.8%) were Gram positive bacteria, mainly streptococci (n=13) and coagulase-negative staphylococci (n=5). Most unidentified streptococci were Streptococcus pneumoniae (eight S. pneumoniae were not identified, score < 1.7, the two remaining S. pneumoniae being identified correctly but with a low score value, i.e. >1.7-2). Moreover, among the 5 Gram negative bacteria with a score below 1.7, 4 were of encapsulated species (2 Klebsiella pneumoniae and 2 Haemophilus influenzae).

Table 1

Identification Number Correct Correct No reliable

identification ¹ identification ² identification

(high score (low score (score <1.7)

>2) 1.7-2)

Gram negative bacilli 46 38 (83%) 3 (7%) 5 (11%)

Escherichia coli 15 14 1

Klebsiella pneumoniae 9 6 1 2

Klebsiella oxytoca 4 3 1

Pseudomonas aeruginosa 4 4

Enterobacter cloacae 3 3

Citrobacter koseri 2 2

Haemophilus influenzae 2 2

Morganella morganii 2 2

Bacteroides distasonis 1 1

Citrobacter freundii 1 1

Fusobacterium necrophorum 1 1

Proteus vulgaris 1 1

Serratia marcescens 1 1

Gram positive cocci 74 31 (42%) 23 (31%) 20 (27%)

Staphylococcus aureus 25 20 5

Staphylococcus epidermidis 23 6 13 4

Streptococcus pneumoniae 10 2 8

Streptococcus agalactiae 5 2 1 2

Staphylococcus hominis 2 2

Streptococcus dysgalactiae 2 2 Enterococcus faecalis 1

Finegoldia magna 1 Micromonas micros 1

Staphylococcus caprae

Staphylococcus haemolyticus 1

Streptococcus bovis 1

Streptococcus pyogenes 1

Others

Brevibacterium casei 2 1 1

Total 122 69 (57%) 27 (22%) 26 (21%)

Correct identification at the species level.

Correct identification at the species level for all but one, for which MALDI-TOF MS identification was congruent with conventional identification at the genus level only

(identified as S. caprae using Vitek identification and S. pasteuri using MALDI-TOF MS identification).

Our results showed that the methods of the invention allow efficient diagnosis of bloodstream infections using MALDI-TOF mass spectrometry. Indeed, identification was obtained for 78.7% of the blood culture pellets analyzed. Moreover, 100% of MALDI-TOF MS identifications were congruent at the species level when considering as valid only score values greater than 2, as proposed by the manufacturer. In addition, 99% of MALDI-TOF MS identifications were matching at the species level with conventional identification when considering both score values of 1.7 to 2 and greater than 2. This excellent performance of coupled ammonium chloride lysis procedure and MALDI-TOF MS analysis was unexpected, since in a recent study comparing mass spectrometry identification of routine bacterial strains with conventional identification, only 84.1% of strains were correctly identified (5). The better result we observed is likely due to the different studied setting, i.e. blood cultures versus a large variety of clinical samples. The lower yield of valid MALDI-TOF MS results with streptococci and staphylococci might be due to the cell wall composition of Gram positive bacteria. In addition, the presence of a capsule may likely also explain the low identification rate of S. pneumoniae, H. influenzae and K. pneumoniae. The use of an ammonium chloride-driven hemolysis before analyzing positive blood cultures by MALDI-TOF MS is a very promising new method allowing fast, accurate and inexpensive identification of the etiological agents of life-threatening bloodstream infections. Example 3: Follow-up and confirmation of results of Example 2

A follow-up study, with the same design, was conducted on 314 monobacterial positive blood cultures bottles (see Example 1).

A valid score was obtained in 59% of cases, a low score in 28% and no reliable identification in 13% of cases. Seven discordant (2.2%) results were observed between Maldi-tof and conventional identification. Six were with a valid score (>=2), and one with a low score (1.7- 2). Table 2 details the discordances in brackets.

Table 2

Two discordances were observed for gram negative bacilli, the identifications being correct at genus level only. One Enterobacter cloacae was identified as Enterobacter species, one Pseudomonas putida as Pseudomonas koreensis.

Five discordances occurred for gram positive bacteria, in 4 cases the identification were correct at genus level. In 3 cases, Streptococcus mitis were misidentified as Streptococcus pneumoniae by Maldi-tof with a valid score and in one case, Staphylococcus epidermidis gave Staphylococcus capitis by Maldi-tof.

In one case, a misidentification at genus level occurred, the Maldi-tof giving Staphylococcus epidermidis with a low score instead of Streptococcus pneumoniae.

Maldi-tof MS direct identification was obtained for 85% of the pellets in the follow-up study compared to 79% in the initial study. 98% of Maldi-tof direct identifications were congruent at species level with conventional identifications done on colonies compared to 99% in the initial study.

The follow-up study confirms the excellent results obtain with the simple ammonium chloride based pellet preparation from positive blood culture.

Bacterial pellet obtained with the ammonium chloride procedure was also tested for other diagnostic methods than MALDI-TOF. The following study report the results obtained with the automate Vitek (bioMerieux) used for both the identification and susceptibility testing.

Example 4: Antimicrobial susceptibility testing using the bacterial preparation

Objectives: Blood cultures are crucial to establish the etiology of bloodstream infections. Rapid identification and antimicrobial susceptibility tests (AST) are pivotal to guide antibiotherapy. This may be accurately done using the Vitek 2 system inoculated with subcultures. Direct testing of bacterial pellets from positive blood cultures is unsatisfactory. Thus, we developed a procedure to prepare pellets that may be directly used to identify both Gram-positive and Gram- negative bacteria with Vitek cards. Methods: Pellets from positive blood culture vials (Bactec 9240) were treated with ammonium chloride lysing solution and centrifuged. GN and GP Identification Vitek cards were used for Gram-positive and Gram- negative bacteria identification, respectively. The AST-P580 and AST-GN26 Vitek cards were used for AST of staphylococci and Enterobacteriaceae, respectively. Identification and AST were repeated using subcultures (gold standard). Identification was classified as correct, misidentification and non- identification. For AST, discordance results were classified as: very major (false susceptible); major (false resistant) and minor errors (all others).

Results: Overall 232/291 blood cultures (80%) gave correct identification at the species level by Vitek direct inoculation. The identification were correct for 89 Gram-negative bacteria (98%), 129 staphylococci (72%), 10 enterococci (91%), 4 streptococci (40%). Misidentification occurred for 32 bacterial pellets, 0 for Gram-negative bacteria, 31 for staphylococci (5 at genus level), 0 for enterococci, 1 for streptococci. AST results were analysed for 212 isolates with congruent identification. For 83 Enterobacteriaceae, the overall agreement was 98%. The number of very major, major, minor errors were 1 (0.1%), 6 (0.5%), 30 (2.4%), respectively. For 129 staphylococci, the overall agreement was 96%. The number of very major, major, minor errors were 24 (1.6%), 3 (0.2%), 75 (4.8%), respectively.

Conclusion: Bacterial pellets from positive blood cultures prepared with an ammonium chloride-driven hemolysis allow direct inoculation of Vitek cards used for identification and for antimicrobial susceptibility testing with excellent accuracy for Enterobacteriacae and lower accuracy for staphylococci.

References:

1. Friedrichs, C, A. C. Rodloff, G. S. Chhatwal, W. Schellenberger, and K. Eschrich. 2007. Rapid identification of viridans streptococci by mass spectrometric discrimination. J

Clin Microbiol 45:2392-7.

2. Smole, S. C, L. A. King, P. E. Leopold, and R. D. Arbeit. 2002. Sample preparation of Gram-positive bacteria for identification by matrix assisted laser desorption/ionization time-of-flight. J Microbiol Methods 48: 107-15.

3. Conway, G. C, S. C. Smole, D. A. Sarracino, R. D. Arbeit, and P. E. Leopold. 2001. Phyloproteomics: species identification of Enterobacteriaceae using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J Mol Microbiol Biotechnol 3: 103-12.

4. Mellmann, A., J. Cloud, T. Maier, U. Keckevoet, I. Ramminger, P. Iwen, J. Dunn, G. Hall, D. Wilson, P. Lasala, M. Kostrzewa, and D. Harmsen. 2008. Evaluation of matrix- assisted laser desorption ionization-time-of-flight mass spectrometry in comparison to 16S rRNA gene sequencing for species identification of nonfermenting bacteria. J Clin Microbiol 46: 1946-54.

5. Seng, P., M. Drancourt, F. Gouriet, B. La Scola, P. E. Fournier, J. M. Rolain, and D. Raoult. 2009. Ongoing revolution in bacteriology: routine identification of bacteria by matrix - assisted laser desorption ionization time-of-flight mass spectrometry. Clin Infect Dis 49:543- 51.