WO2015121784A1

WO2015121784A1 - Method for detection of anti-rituximab antibodies

Info

Publication number: WO2015121784A1
Application number: PCT/IB2015/050949
Authority: WO
Inventors: Ravi Kumar LELLA; Narendra Babu MORAM; Sridevi KHAMBHAMPATY
Original assignee: Dr. Reddy's Laboratories Limited
Priority date: 2014-02-11
Filing date: 2015-02-09
Publication date: 2015-08-20

Abstract

This invention is related to a method for detecting anti-rituximab antibodies in a RF positive plasma samples, wherein the method mitigated the interference of RF to a significant level using biotinylated F(ab)₂ fragments of rituximab as detecting agent. The method also used a pre-treatment step with acidic buffer containing glycine, prior to subjecting the samples for detection of anti-rituximab antibodies.

Description

METHOD FOR DETECTION OF ANTI-RITUXIMAB ANTIBODIES

RELATED APPLICATION

This application is related to and takes priority from Indian Provisional Application 644/CHE/2014 filed 1 1 ^th February, 2014 and is herein incorporated in its entirety.

BACK GROUND

Protein based therapeutics, in particular monoclonal antibodies, are highly successful in treating various diseases including oncological disorders. However, when administered for treatment, these therapeutic drugs often induce undesirable immune response in patients. One such common drug induced adverse response is the production of anti-drug antibodies against the therapeutic antibody in use.

Anti-drug antibodies (ADA) are known to form immune complexes with the drug, leading to drug clearances and low drug levels in the system. This severely affects the clinical efficacy of the drug. Further, complexes formed by ADA may result in increased hypersensitivity reactions impacting the safety of the drug. As a consequence, significant alterations are seen in the pharmacokinetic and

pharmocodynamic property of the administered drug. Thus screening and detection of ADA becomes highly important, and is mandated by the regulatory agencies when assessing the safety/immunogenicity of a therapeutic drug.

(http://www.fda.gov/downloads/Drugs/.../Guidances/UCM192750.pdf)

Most assay methodologies used for the detection and quantification of ADA depend on the specific binding of the ADA to its target drug via the antigen binding site. However, these binding sites are often partially or completely masked by multiple factors that also bind to the ADA, which then prevents the latter's binding to the target drug. This in turn results in underestimation of the level of ADA present in a sample causing interference in the performance or sensitivity of the assay. Such factors differ depending up on the disease and the nature of the therapeutic drug molecule. For eg., in case of rheumatoid arthritis (RA), serum or plasma from patients most often contain high concentrations of rheumatoid factors (RF) which are antibodies with high affinity to the Fc region of IgG that can complex with ADA, causing interference to the ADA detection assay. Further, therapeutic monoclonal antibodies are generally administered at higher doses, and possess relatively longer half-lives. This results in the presence of higher levels of the monoclonal antibodies in blood, which can also interfere with the detection assay by forming immune complexes with the ADA. Thus any assay employed to detect ADA against a specific drug requires customization and optimization for reducing such interfering factors.

Rituximab, a chimeric anti-CD20 monoclonal antibody widely used for the treatment of RA, often results in the production of anti-rituximab antibodies as an adverse immune response in individuals. As described above, high concentrations of RF and rituximab present in RA positive plasma or serum sample may strongly interfere in the detection of anti-rituximab antibodies.

Hence, the primary objective of the invention is to develop an immunoassay for the detection of anti-rituximab antibodies in which the interference from RF is significantly reduced or tolerated. Another object of the invention is to measure anti- rituximab antibodies in presence of excess free drug viz., rituximab, wherein the interference from the drug is reduced or tolerated. The present invention discloses a method wherein interferences from RF and the drug is tolerated or reduced significantly such that the interfering factors have little impact on the sensitivity or performance of the assay.

SUMMARY

The present invention discloses a method for detecting anti-rituximab antibodies in RF positive RA patient's plasma sample using labelled F(ab)₂ fragments of rituximab as a detecting agent in an enzyme linked immunosorbent assay (ELISA). Use of the F(ab)₂ fragment of rituximab, instead of whole antibody, mitigates RF interference in the assay since, the said fragment lacks the 'Fc' portion of the antibody that is specifically required for the binding of RF.

Further, prior to detection, the method employs acid treatment of sample with a buffer containing glycine for dissociation of immune complexes (present/formed in the sample) and retrieval of anti-rituximab antibodies. The acid treatment along with the use of 'fragmented antibody' mitigates RF interference substantially, to a level such that the kind and degree of ADA detection in RF positive plasma sample (with high concentration of RF) is comparable or similar to that of the level obtained in normal human plasma sample (with no or a little RF). This in turn indicates that the RF present in high concentrations (~1500 units/ml) in the sample is either removed or does not impact the performance of the assay.

Additionally, the method can detect ADA in presence of rituximab (free drug) as high as 500 μg/ml of the plasma sample, indicating high drug tolerance of the assay. BRIEF DESCRIPTION OF DRAWINGS

Figure 1 : Illustrates a comparison between NPHP and RAPHP samples (spiked with anti-rituximab antibodies), wherein the detection agent is biotinylated whole rituximab. Data for both, acetic acid and glycine treated samples were shown.

Figure 2: Illustrates a comparison between NPHP and RAPHP samples (spiked with anti-rituximab antibodies), wherein the detection agent is biotinylated F(ab)2 fragments of rituximab. Data for both, acetic acid and glycine treated samples were shown.

Figure 3: Illustrates the sensitivity of the assay in presence of varying concentrations of rituximab. DETAILED DESCRIPTION

Detection of anti-rituximab antibodies (ARiA) in a RF positive Rheumatoid Arthritis (RA) patient sample is complicated by the presence of high concentrations of RF and free drug i.e., rituximab, which interferes with the sensitivity and

performance of the assay. These interferences contribute to false positive/negative results often causing improper estimation of anti-rituximab antibody level. The method disclosed in the current invention is able to detect ARiA without any interferences from RF and/or free rituximab.

RF has high affinity towards Fc portion of IgG antibodies, and in turn to ARiA which is an IgG molecule resulting in the formation of RF-ARiA complex. Additionally free rituximab found in excess in RA treated patients' sample, having specific affinity towards ARiA tend to form complexes with ARiA. These complexes block the binding sites of ARiA preventing it from binding to the target detecting agent and subsequently interfering in its detection. Hence, it is essential to dissociate these complexes (RF-ARiA and Rituximab-ARiA and RF-rituximab) present or formed in a sample, prior to subjecting to the ARiA detection.

The present invention discloses a method wherein an acidic buffer containing glycine is used for dissociating the immune complexes and the conditions are optimized to result in complete dissociation of the complexes. In addition, the method employs the acidic buffer that is used for dissociation, for retrieval of ARiA from an affinity capture matrix, which buffer prevents formation of further complexes with the affinity agent. The method also uses intermittent wash steps to remove/reduce the concentrations of free RF and rituximab obtained during the dissociation and retrieval process. The retrieved ARiA that is significantly free from RF and/or rituximab is detected using the F(ab)2 fragment of rituximab which specifically binds the ARiA molecule. Minor amounts of RF even if present, does not bind to the detecting F(ab)₂ fragment (as the fragment lacks the Fc portion of the whole antibody) thereby removing/reducing any reactive signal from RF.

The results obtained reveal that the method remove/tolerate RF

concentrations as high as ~ 1500 units /ml of the sample. The RF tolerance level obtained using the method is noteworthy, since the RF concentration present in established diseased conditions of RA is greater than 20 units/ml and generally an average of about 78 units /ml of the plasma sample {Rehnberg et. al, Arthritis Research and Therapy 2009, volume 11, issue (4)). The method detects ADA in high concentrations of RF without any interference from the factors, demonstrate its utility in highly progressed diseased condition of RA.

In addition, the method is able to detect ADA in the presence of high concentrations of the drug, rituximab, that it can detect about 100 ng/ml of ARiA in the presence of 100 μg/ml of rituximab. The results accordingly indicate the application of the method in the clinical and mimic near real life scenario of diseased condition. Various embodiments disclosed herein as methods describe the invention in detail.

In an embodiment, the claimed invention discloses a method of detecting anti- rituximab antibodies in a RF positive sample using enzyme linked immune sorbent assay, wherein labelled F(ab)₂ fragment of rituximab is used as the detecting agent and wherein the said method mitigates RF interference substantially.

In an embodiment, the claimed invention discloses a method of detecting anti- rituximab antibodies in a RF positive sample comprising the steps of; a) incubating the sample with an acidic buffer containing glycine to dissociate immune complexes

b) retrieving the anti-rituximab antibodies from the said sample using the acidic buffer of step a)

c) coating the retrieved ARiA onto a solid phase

d) adding a labelled F(ab)₂ fragment of rituximab to the said solid phase to bind ARiA and

e) detecting the said labelled fragment and in turn the bound ARiA, using

standard protocols of enzyme linked immunosorbent assay methodology, wherein, the said method mitigates RF interference substantially.

In another embodiment, the said acidic buffer has pH values in the range of 2.0 to 3.0. Preferably, the pH of the acidic buffer is 2.5.

In yet another embodiment of the invention, the concentration of glycine in the said acidic buffer is from about 0.2 M to 0.4 M.

In yet another embodiment of the invention, the said labelled F(ab)₂ fragment of rituximab is biotinylated F(ab)₂ fragment of rituximab.

The biotinylated F(ab)₂ fragment of rituximab is detected using the principle of enzyme-substrate colorimetric reaction wherein, streptavidin coupled horseradish peroxidase is used as an enzyme and 3,3',5,5'-Tetramethylbenzidine (TMB) as a substrate. However, the F(ab)₂ fragment can be labelled with any commercially available suitable labelling agent (such as but not limited to ruthenium, iodine etc.,) and shall be detected using appropriate substrate using the principle of enzyme linked immunosorbent assay.

The method disclosed in the invention, can detect ARiA as low as 4 ng/ml of the sample in the absence of rituximab. In yet another embodiment the present invention is used to assay anti-drug antibodies in sample that contain RF interferences.

In yet another embodiment, the present invention is used as a sensitive antidrug antibody assay in sample containing RF interference, wherein the drug is a chimeric or murine therapeutic antibodies. The method as disclosed in the invention can detect ARiA in the presence of excess of free rituximab viz., as high as 500 μg of rituximab /ml of the sample.

Further, the method is sensitive such that, it can detect on an average about 100 ng/ml of ARiA (and as low as 62.5 ng/ml of ARiA) in presence of 100 μg/ml of rituximab. "Substantially" refers to the performance of the assay wherein, when detection of ARiA is performed in RF positive plasma and normal human plasma sample, the results (in terms of mean absorbance value and P/N ratio) obtained in both the samples are similar or comparable. 'Substantially' also refers to a level wherein the RF factors as high as about 1500 units/ml present in the sample, is

removed/reduced to a level such that it does not impact or interfere in the detection of ARiA.

The term 'sample' as used herein the invention refers to an RF positive plasma or serum sample obtained from patients who are not treated with rituximab and, in which case, the sample obtained is spiked with anti-rituximab antibodies and/or rituximab. However, the method as disclosed in the invention can be used for the detection of ARiA in RF positive plasma/serum sample obtained from patients treated with rituximab.

The term "anti-rituximab antibody (ARiA)" used here in refers to an antibody that is raised/formed/produced against rituximab in animals including humans. "P/N ratio" is a ratio between mean absorbance of positive control to mean absorbance of negative control. The ratio helps in normalization of data as it dissociates the signal from the background noise. This ratio is also known as highest signal to noise ratio. "Positive control" as used here in the invention, is a sample (plasma / serum from RF positive RA patients or plasma/serum from healthy volunteers) spiked with anti-rituximab antibody. Some of the positive controls are spiked with rituximab to mimic the physiological condition of rituximab treated patient's sample and/or to evaluate drug tolerance. The term 'drug tolerance' as is defined as the maximal amount of free drug i.e., rituximab, present in the sample that still results in detectable ARiA signal.

Hence, to determine the drug tolerance, samples from RF positive patients (not treated with rituximab) are spiked with different concentrations of rituximab and anti- rituximab antibodies. "Negative quality control (NQC)" is an unspiked sample (plasma / serum from

RF positive RA patients or plasma/serum from healthy volunteers) in which no anti- rituximab antibodies or rituximab have been added.

"Sensitivity" of the assay is defined as the lowest concentration of positive control antibody preparation which consistently provides signal in the assay. "RAPHP" is the rheumatoid factor positive pooled human plasma samples from rheumatoid arthritis patients obtained from authorized blood banks.

"NPHP" is normal pooled human plasma samples from healthy volunteers obtained from authorized blood banks.

Certain specific aspects and embodiments of the invention are more fully described by reference to the following examples. However, these examples should not be construed as limiting the scope of the invention in any manner. EXAMPLES

During the development of the assay for detection of anti-rituximab antibodies various parameters were optimized. One such parameter was evaluation of

dissociation and elution capacity of two different acidic buffers. Two different acidic buffers namely 300 imM acetic acid, pH 2.5 and 0.2 M glycine at pH 2.5 were tried and evaluated for the dissociation of immune complexes, that were formed in the samples. The method was also evaluated for two detection agents such as biotinylated rituximab (whole antibody) and biotinylated F(ab)2 fragments of rituximab to assess the mitigation of interferences in the assay. Further, to determine the performance of the assay, the method was executed in plasma samples obtained from both, normal healthy individuals and RA patients and the results were compared.

Example 1 : Sample preparation

Plasma sample obtained from various rheumatoid factor positive RA patients (not treated with rituximab) were pooled and spiked with rat-anti-rituximab antibodies (positive control - Rheumatoid Arthritis Pooled Human Plasma (RAPHP)). To compare the study with normal samples, plasma samples were collected from normal healthy individuals, pooled, and spiked with rat-anti-rituximab antibodies (positive control - Normal Pooled Human Plasma NPHP). Further, some of the positive controls were spiked with rituximab to mimic physiological conditions of rituximab treated (drug administered) RA patient and to determine the drug tolerance level of the assay.

Negative quality control samples were unspiked plasma samples that are pooled plasma samples either from RA patients or normal healthy individuals.

RAPHP used in all experiments were estimated to contain not less than 1 500 units of rheumatoid factors/ml of sample.

The spiked and unspiked samples were incubated at 37 ^QC for 1 h under shaking conditions to allow the formation of immune complexes. Samples were diluted in acidic buffer. Example 2: Preparation of detecting agents a) Biotinylated rituximab:-

Rituximab was added to biotin solution and the mixture was incubated for 30 minutes to allow conjugation of biotin to rituximab. Excess of unreacted biotin was removed using zeba desalt column (size exclusion).

Biotinylated rituximab was then purified using avidin affinity column. b) Generation of biotinylated F(ab)₂ fragments of Rituximab:-

Biotinylated rituximab (Bi-RI) was used for the generation of (Fab)₂ fragments of rituximab. Pepsin was added to Bi-RI and the mixture was incubated at 37^Q C for 4 h, to obtain F(ab)₂ fragment of rituximab following which the reaction was terminated using Tris buffer.

Further, undigested biotinylated rituximab was removed by using protein A solid support and the purified biotinylated F(ab)₂ fragments of rituximab (Bi-F(ab)₂-RI) was obtained. Bi-F(ab)₂-RI was diluted in 1 % BSA for further experiments. Example 3: Acid dissociation of immune complexes

NPHP and RAPHP samples that were spiked with rat-anti rituximab antibodies (positive controls) were incubated at 37 ^QC for 1 h under shaking conditions to allow the formation of immune complexes. Samples were then diluted in acidic buffers, 300 imM acetic acid or 0.2 M glycine, and incubated separately for 30 min at room temperature for dissociating the immune complexes formed in the samples.

Example 4: Retrieval of ARiA

Acid treated samples, i.e., acetic acid or glycine treated, individually were obtained, neutralized with Tris buffer (pH 9.5±0.05) and added to the micro titer plates that were coated with 100 μΙ of rituximab. The two separate samples were then incubated, under shaking conditions, for 1 h at 37 ^QC. The plates were then washed with phosphate buffer saline tween 20 (PBST) to remove unbound material. ARiA bound to rituximab was eluted using 300 imM acetic acid or 0.2 M Glycine buffer at pH 2.5. The same acidic buffer that is used for the dissociation of immune complexes is also used for the retrieval of ARiA.

Example 5: Detection of ARiA

Retrieved ARiAs (either using acetic acid or glycine) from example 4, were transferred to uncoated plates containing of 1 M Tris buffer (pH 9.5±0.05)) and the plates were sealed and incubated for 1 h at 37 ^QC under shaking conditions. The plates were then washed with PBST buffer and blocked using 1 % BSA. Followed by which, the plates were again washed with PBST buffer. ARiAs bound to the plates were then detected either by biotinylated rituximab (whole antibody) or biotinylated F(ab)₂ fragments of rituximab by adding the detecting agents separately to the plates. The plates were incubated at 37 ^QC for 1 h under shaking condition. The plates were then washed with PBST buffer. Streptavidin-Horse Radish Peroxidase (which was diluted in 1 % BSA ) was added to the plates and incubated at 37 ^QC for 1 h under shaking conditions.. The plates were washed with PBST buffer for three times and 100 μΙ of 3,3',5,5'-Tetramethylbenzidine (TMB) substrate was added and incubated at 37 ^QC for 1 h under shaking conditions. 100 μΙ of stop solution was added to stop the reaction and absorbance at 450 nm was measured using a micro plate reader.

Results of the comparative assay are given below in Table 1 and 2 which are also represented as Figure 1 and 2. Data was analyzed in terms of mean absorbance and P/N ratio (positive control/negative quality control).

P/N ratio =

mean absorbance of positive control /

mean absorbance of negative quality control Table 1 : Absorbance and P/N ratios of 300 mM acetic acid (pH 2.5) treated NPHP and RAPHP samples.

Table 2: Absorbance and P/N ratios of 0.2 M glycine treated NPHP and RAPHP samples

From the above results, it was evident that, biotinylated F(ab)₂ fragments showed higher P/N ratios as compared to biotinylated whole antibody molecule. Also, higher P/N ratios were observed with 0. 2 M glycine as compared to 300 mM acetic acid. It is significant to note that, use of glycine treatment and F(ab)2 fragment of rituximab resulted in P/N ratios, that were similar and comparable for RAPHP and NPHP samples, in turn indicating that the RF interference is mitigated substantially.

Example 6: Sensitivity of the assay in presence of varying concentrations of rituximab

Sensitivity of the assay in presence of various concentrations of rituximab was performed by spiking rituximab in samples containing different concentrations of rat- anti rituximab antibodies. Samples were incubated for 2 h followed by acid treatment of samples with 0.2 M Glycine at pH 2.5. Following acid treatment the assay was performed as described in Examples 4 and 5. Results obtained are listed in Table 3 and the same is represented as Figure 3. The screening cut point for this experiment is set to be 1 .189. "Screening Cut point" of the assay is the level of response at or above which a sample is identified to be positive and below to be negative. Table 3: Represents P/N ratios of anti-rituximab antibody in the assay in presence of varying concentrations of rituximab.

Claims

CLAIMS We Claim:

1 . A method for detecting anti-rituximab antibodies (ARiA) in a RF positive sample, wherein the said method mitigates RF interference comprising the steps of; a) incubating the sample with an acidic buffer containing glycine to dissociate immune complexes b) retrieving the anti-rituximab antibodies from the said sample using the acidic buffer of step a) c) coating the retrieved ARiA onto a solid phase d) adding labelled F(ab)₂ fragment of rituximab to the said solid phase to bind ARiA and e) detecting the said labelled fragment and in turn the bound ARiA, using enzyme linked immunosorbent assay.

2. A method according to claim 1 , where in the said acidic buffer has pH values in the range of 2.0 to 3.0.

3. The acidic buffer according to claim 2, is a glycine buffer, wherein the concentration of glycine in the said acidic buffer is about 0.2 M to 0.4 M.

4. A method according to claim 1 , wherein the said labelled F(ab)₂ fragment of rituximab is biotinylated F(ab)₂ fragment of rituximab.

5. A method for detecting anti-drug antibodies (ADA) in a RF positive sample, wherein the said method mitigates RF interference comprising the steps of; a) incubating the sample with an acidic buffer containing glycine to dissociate immune complexes b) retrieving the anti-drug antibodies from the said sample using the acidic buffer of step a) c) coating the retrieved anti-drug antibodies (ADA) onto a solid phase d) adding labelled F(ab)₂ fragment of said drug to the said solid phase to bind antidrug antibodies (ADA) and e) detecting the said labelled fragment and in turn the bound anti-drug antibodies (ADA), using enzyme linked immunosorbent assay.

6. The anti-drug antibodies according to claim 6, wherein the drug is a chimeric or murine therapeutic antibodies