2023 Volume 46 Issue 4 Pages 621-629
Monitoring serum infliximab (INF) concentrations is crucial for designing appropriate doses for patients with rheumatoid arthritis. It is recommended to maintain the serum trough INF level at least 1.0 µg/mL. In Japan, an in vitro diagnostic kit using immunochromatography has been approved to determine whether the serum INF concentration is over 1.0 µg/mL or not, and to support the determination of the necessity of increasing the dose or switching to another drug. Biosimilars (BS) of INF may have immunochemical properties different from those of its innovator product, which may show different reactivities on the diagnostic kit. In this study, the responses of the innovator and five BS products on the kit were compared. Based on visually comparing the intensity of color development between the test and control samples, differences were found in the judgment results depending on the analyst. In particular, 1.0 µg/mL was not determined as positive in some cases, whereas 2.0 µg/mL was reliably determined as positive. Overall, no significant difference in reactivity was found between the innovator and five BS products. To further compare the differences in immunochemical properties, the reactivity of these products with three enzyme-linked immunosorbent assay (ELISA) kits was compared. The results confirmed that there were no significant differences among the innovator and BS products in reactivity with the examined kits. When using that diagnostic kit, the users need to be aware that the judgement around 1.0 µg/mL INF may differ depending on the test conditions, including the analyst.
Therapeutic drug monitoring (TDM) is a useful approach for designing optimal doses for individual patients. In particular, for drugs with a narrow therapeutic window, TDM is crucial to obtain sufficient therapeutic effects and reduce adverse events. In addition to drug concentrations, pharmacodynamic markers and other biomarkers have also been used as monitoring parameters for TDM. Drugs subject to TDM at Japanese hospitals generally include antiepileptics, anti-arrhythmia, and immunosuppressive drugs; aminoglycoside and glycopeptide antibiotics; and triazole antifungal agents. In vitro diagnostic kits based on various methods, including immunochemical methods, have been approved for detecting many of these chemical drugs.
Along with chemical drugs, TDM has also been used for biopharmaceuticals. The role of TDM has been increasingly emphasized, particularly in immune-mediated inflammatory diseases, including rheumatoid arthritis (RA) and inflammatory bowel.1,2) This is because anti-tumor necrosis factor α (TNFα) drugs, such as infliximab (INF), adalimumab, and golimumab, show higher therapeutic effects, whereas some patients exhibit primary no response or secondary loss of response to these drugs. In the case of INF for treating RA, it has been clarified that decreased disease activity is correlated with increased trough serum concentrations of INF (ATTRACT and START studies).3,4) The ATTRACT study involving INF co-administration with methotrexate in patients with RA indicated that a serum trough INF concentration of ≥1.0 µg/mL is needed to achieve efficacy. The START study showed that low trough INF levels (< 1.0 µg/mL) are associated with the need for dose escalation to improve efficacy. Another report also indicated that the magnitudes of clinical responses and progress inhibition of joint destruction were correlated with trough serum INF levels.5) In this clinical study, the median serum trough levels in good and moderate responders were 1.1 µg/mL. These studies indicate that the threshold INF serum concentration required to achieve remission in rheumatoid arthritis (RA) is approximately 1.0 µg/mL. Therefore, serum trough INF concentration is a useful index to be monitored for optimizing treatment strategies and is recommended to be maintained at least at 1.0 µg/mL. Notably, in a retrospective cohort study, monitoring the serum INF level is especially important for avoiding the generation of second non-responders, who gradually stop responding to INF during treatment.6)
In this context, an in vitro diagnostic kit based on immunochromatography was approved in Japan to determine whether the serum INF concentration exceeded 1.0 µg/mL, and to support the determination of the necessity of dose increase or change to another drug.7,8) The result can be obtained in approximately 20 min and used to alter the treatment strategy during outpatient consultation. As of October 2022, five products of INF biosimilars have been approved in Japan. However, the effectiveness of this diagnostic kit was confirmed only with the innovator INF, and its applicability to biosimilars of INF remained unclear.7) Owing to expectations of reduced pharmaceutical costs, biosimilar development has been pushed for the past few years.9) Biosimilars may have immunochemical properties different from those of the innovator because of differences in the manufacturing process, including cell substrates. Although the quality attributes of biosimilars are highly similar to those of the innovator product, which ensure their efficacy and safety, the reactivity of these products with binding reagents in diagnostic kits is not considered in the evaluation of their quality attributes during development. Thus, INF biosimilars may exhibit different results when analyzed with the diagnostic kit and are thus not mentioned in the intended use of the kit.
In this study, the responses of the innovator and biosimilars of INF on that diagnostic kit were compared. Recently, one of the INF biosimilar products was reported to show an equivalent result to the innovator using this diagnostic kit.10) We compared the responses of all INF products available in Japan, i.e., the innovator and five biosimilar products, on the diagnostic kit.
The INF products evaluated in this study are listed in Table 1. All INF products, Remi-check Q (LSI Medience Corporation, Tokyo, Japan), and OTSUKA distilled water (Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan) were purchased from commercial suppliers. Three lots (#B804, #R905, and#B007) of Remi-check Q were used. Three enzyme-linked immunosorbent assay (ELISA) kits for measuring serum INF concentration, Infliximab (Remicade) ELISA SHIKARI Q-INFLIX (Matrix Biotechnology Co., Ltd., Ankara, Turkey), IDKmonitor Infliximab drug level ELISA (Immunodiagnostik AG, Bensheim, Germany), and ImmunoGuide Infliximab ELISA (TaniMedical Ltd., Ankara, Turkey) were used.
| Category | Product name | Abbreviation | Lot No. | Host cell | Additives in drug product | pH after dissolving | Manufacturer |
|---|---|---|---|---|---|---|---|
| Innovator | REMICADE™ for I.V. Infusion | Remicade | JAL56012 | Sp2/0 | Sucrose Polysorbate 80 Sodium dihydrogen phosphate monohydrate Disodium hydrogen phosphate dihydrate | 6.9–7.5 | Janssen Pharmaceutical K.K./ Mitsubishi Tanabe Pharma Corporation |
| Biosimilar 1 | Infliximab BS for I.V. Infusion 100 mg [NK] | BS1[NK] | 451001 | Sp2/0 | Same as above | 6.9–7.5 | Nippon Kayaku Co., Ltd. |
| Biosimilar 1 | Infliximab BS for I.V. Infusion 100 mg [CTH] | BS1[CTH] | 17CA01 | Sp2/0 | Same as above | 6.9–7.5 | CELLTRION Healthcare Japan Inc. |
| Biosimilar 2 | Infliximab BS for I.V. Infusion 100 mg [AYUMI] | BS2[AYUMI] | C00300 | CHO | Same as above | 6.9–7.5 | AYUMI Pharmaceutical Corporation |
| Biosimilar 2 | Infliximab BS for I.V. Infusion 100 mg “Nichiiko” | BS2[Nichiiko] | D00200 | CHO | Same as above | 6.9–7.5 | Nichi-Iko Pharmaceutical Co., Ltd. |
| Biosimilar 3 | Infliximab BS for I.V. Infusion 100 mg [Pfizer] | BS3[Pfizer] | Y01110 | CHO | Sucrose Polysorbate 80 Succinic acid Disodium succinate hexahydrate | 5.5–6.5 | Pfizer Japan Inc. |
The innovator and five INF biosimilars were reconstituted according to their package inserts. Freeze-dried drug preparations (100 mg of INF per vial) were dissolved in 10 mL of distilled water with gentle mixing. The protein concentration of each solution was measured using a NanoDrop 2000c spectrophotometer (Thermo Scientific, DE, U.S.A.). The solution was then stored in small aliquots at −80 °C. The stock solutions were diluted to 1 mg/mL with pooled human serum (CT-0365964, Clinical Trials Laboratory Service Limited (London, U.K.); hereafter abbreviated as human serum 1) before use. The solutions were further diluted with the same pooled serum, and 0.3, 0.7, 1.0, 1.3, and 2.0 µg/mL test samples were prepared.
Measurement with the Infliximab Detection Kit, Remi-Check QThe measurement procedure was performed according to the manufacturer’s instructions in the package insert of Remi-Check Q (Q 2020) with some modifications.7) For triplicate measurements, 80 µL of each test sample was diluted with 320 µL of sample diluent. The diluted sample (120 µL) was then dropped onto the application field of the test plate. Then, the same volume of the cut-off control was immediately dropped onto the designated field. After leaving the plate in a horizontal place for 15 min, the intensity of red-purple coloration of the respective lines in the results field was determined visually. In the experiments shown in Table 2, data on 0.3 µg/mL of BS2[Nichiiko] were obtained in duplicate due to the limited number of the kit (Lot#B804) that were available for this study. In cases where the color intensity of the test sample line was equal to or higher than that of cut-off, the result was determined as positive (> 1 µg/mL). If the coloration of the test sample line was less than that of the cutoff, the result was determined as negative (less than 1 µg/mL). Determination by visual inspection was performed by more than three analysts in an independent and blinded manner. All the kit plates tested were photographed for recording.
| Sample | Analyst | Infliximab concentration (µg/mL) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.3 | 0.7 | 1.0 | 1.3 | 2.0 | ||||||||||||
| 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | ||
| Innovator | A | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + |
| B | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| D | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| E | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| BS1[NK] | A | — | — | — | — | — | — | — | — | — | — | + | — | + | + | + |
| B | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | — | — | — | — | + | + | + | |
| D | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| E | — | — | — | — | — | — | — | — | + | + | + | + | + | + | + | |
| BS1[CTH] | A | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + |
| B | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + | |
| D | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + | |
| E | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| BS2[AYUMI] | A | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + |
| B | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| D | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + | |
| E | — | — | — | — | — | — | + | — | — | + | + | + | + | + | + | |
| BS2[Nichiiko] | A | — | — | — | — | — | — | — | — | — | + | — | + | + | + | |
| B | — | — | — | — | — | — | — | — | — | + | + | + | + | + | ||
| C | — | — | — | — | — | + | + | + | + | + | + | + | + | + | ||
| D | — | — | — | — | — | — | — | — | — | + | — | + | + | + | ||
| E | — | — | — | — | — | — | — | — | — | + | — | + | + | + | ||
| BS3[Pfizer] | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| D | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + | |
| E | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
Infliximab detecting kit: Lot#B804, Diluent: human serum 1. Data on 0.3 µg/mL of BS2[Nichiiko] were obtained in duplicate due to the limited number of the kit (Lot#B804) that were available for this study.
The stock solutions of Remicade and BS3[Pfizer] were dialyzed with three different types of buffers (phosphate buffer (pH 7.2), phosphate buffer (pH 6.0), and succinate buffer (pH 6.0)) using a dialysis kit (Thermo Slide-A-Lyzer MINI Dialysis Unit, 10 K MWCO#69570, Pierce Biotechnology (IL, U.S.A.)) at 4 °C overnight. The protein concentrations of the solutions collected from the dialysis devices were measured and diluted to 1 mg/mL with human serum 1. As a reference, the stock solutions of the innovator and BS3 [Pfizer] were also diluted to 1 mg/mL with human serum 1. These samples were further diluted with the human serum 1, to prepare 0.3, 0.7, 1.0, 1.3, and 2.0 µg/mL concentrations for each test sample.
Effect of Serum Used for Sample Dilution on Background ColorationTo investigate the effects of serum, in addition to human serum 1, another pooled human serum (SER019A050D001, BIOPREDIC International (Saint-Gregoire, France), hereafter human serum 2) was used to dilute the INF stock solution. For comparison, bovine serum albumin (BSA) in phosphate buffered saline (PBS) was also used as a reference for the results of performance evaluation as per the kit’s package insert (ver. 1).
Immunochemical Responses on ELISA KitsTo confirm the similarity of immunochemical properties between the innovator and five biosimilar (BS) products, three commercially available ELISA kits for measuring serum INF concentrations were used, and the concentration-responses of the standard solutions included in the kit, as well as the in-house INF standard solution on the kits, were measured. Considering the measured concentration-response curves, range, and minimum required dilution specified by the kit instructions, the sample concentration ranges to be tested were determined as 2, 6, and 20 µg/mL. The solutions (1 mg/mL) prepared by diluting the stock solutions of the innovator and five INF BS products with human serum 1 were further diluted using the serum to 2, 6, and 20 µg/mL. The WHO International Standard of INF (NIBSC code 16/170) was also diluted with serum to the concentrations as described above. The concentration of samples diluted with the diluent buffer associated with each kit were measured based on absorbance at 450 nm, following the manufacturer’s instructions.
As several reports have indicated that the efficacy of INF is clearly correlated with its serum trough levels, serum INF concentration is a valuable objective index for optimizing treatment, in addition to optimizing individual treatment plans. The in vitro diagnostic kit for measuring serum INF concentrations can be used to easily and quickly identify patients with RA having a serum INF concentration of less than 1 µg/mL and to assist in the decision of increasing the dose. Currently, this kit cannot be used for patients receiving INF BS products because its package insert clearly states that its efficacy has not been sufficiently confirmed when similar therapeutic antibody products, including INF BS products, are used as test samples. However, the approval and use of BS antibody products is currently expanding.
The quality of BS products has been proven to be highly similar to that of innovator products; however, the immunochemical properties related to reactivity with binding reagents in diagnostic kits may differ between the innovator and BS products. Therefore, in this study, we compared the response of innovator and five BS products of INF to investigate whether the diagnostic kit demonstrates the same performance in determining the serum INF levels in patients treated with INF innovator and BS products. The package insert of the Remi-check Q diagnostic kit defines its detection performance as follows: 0.7 µg/mL of the control yields a negative (−) whereas 1.3 µg/mL yields a positive (+) result. Thus, five INF concentrations (0.3, 0.7, 1.0, 1.3, and 2.0 µg/mL) were chosen to evaluate the reactivity of the innovator and BS products on the kit.
The diluted samples of the innovator and BS products shown in Table 1 were tested using the diagnostic kit (Lot#B804). Table 2 shows the results of the test with the innovator and five BS products spiked in human serum 1 as test samples at the indicated concentrations (n = 3). Visual inspection of coloration obtained with each kit was performed by five analysts (Analysts A–E) in an independent and blinded manner. Four of five analysts determined the results of the 1.0 µg/mL innovator sample on all test plates (n = 3) as negative (−). Regarding the 1.0 µg/mL samples of BS products, four of five analysts determined the results as negative (−) with all the test plates (n = 3) except for BS3[Pfizer]. Even with the 1.3 µg/mL sample, some tested plates were determined as negative, except for BS3[Pfizer]. However, the concentration of 2.0 µg/mL was determined as positive (+) in all test plates by all analysts. In contrast, there were no cases wherein the 0.3 and 0.7 µg/mL samples were determined as positive (+). Owing to the measurement principle of visually comparing the color intensity between the cut-off control and test samples, the variability in results was observed depending on the analyst. On comparing the reactivity of the INF innovator and BS products on the kit, no noticeable differences were observed except for BS3[Pfizer], because the responses of these products (i.e., innovator, BS1[NK], BS1[CTH], BS2[AYUMI], BS2[Nichiiko]) were determined as negative at 0.7 µg/mL, positive at 2.0 µg/mL, and variable at 1.0 and 1.3 µg/mL.
As shown in Table 2, BS3[Pfizer] showed different responses on the kits compared to those of the other INF products, including the innovator, possibly because of the following two factors: differences in the immunochemical properties of the product compared with the others and differences in the additives of the drug product. As described in the product labels, the innovator, BS1[NK], BS1[CTH], BS2[AYUMI], and BS2[Nichiiko] contain sodium dihydrogen phosphate monohydrate and disodium hydrogen phosphate dihydrate as additives, and their pH after dissolving in 10 mL of water for is 6.9–7.5. BS3[Pfizer] contains succinic acid and disodium succinate hexahydrate, and its pH after dissolution in 10 mL of water is 5.5–6.5. Therefore, the effects of different additives and pH values after dissolution in water were investigated further. The stock solutions of innovator INF and BS3[Pfizer] were substituted with three different buffers, pH 7.2 phosphate buffer, pH 6.0 phosphate buffer, and pH 6.0 succinate buffer using dialysis. The original stock solutions and substituted solutions were diluted with pooled human serum and tested using the kit (Lot#B007). The test results are listed in Table 3. The pH of the phosphate buffer did not affect the results. In some cases, the sample (0.7 µg/mL) substituted with succinate buffer was determined as positive (+). The pH and additive differences showed no apparent effects on the test results. Further, there were no notable differences in sample reactivity without dialysis between the innovator product and BS3[Pfizer], which means that the results shown in Table 2 were not reproduced when kits from a different lot were used. Therefore, the reason for the difference in reactivity only for BS3[Pfizer] in the first examination remains unclear; however, lot-to-lot variability of the kit might have affected the results.
| Sample | Dialysis | Analyst | Infliximab concentration (µg/mL) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.3 | 0.7 | 1.0 | 1.3 | 2.0 | |||||||||||||
| 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | |||
| Innovator | No dialysis | A | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | — | + | + | + | + | + | + | + | + | ||
| Phosphate buffer (pH 7.2) | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| B | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | ||
| Phosphate buffer (pH 6.0) | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| B | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + | ||
| Succinate buffer (pH 6.0) | A | — | — | — | + | — | — | + | + | + | + | + | + | + | + | + | |
| B | — | — | — | + | — | — | + | + | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | + | + | + | + | + | + | + | + | + | + | ||
| BS3[Pfizer] | No dialysis | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | ||
| Phosphate buffer (pH 7.2) | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| B | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + | ||
| Phosphate buffer (pH 7.2) | A | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | |
| B | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | — | + | — | + | + | + | + | + | + | ||
| Succinate buffer (pH 6.0) | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + | |
| B | — | — | — | + | — | — | — | + | + | + | + | + | + | + | + | ||
| C | — | — | — | — | — | — | — | + | + | + | + | + | + | + | + | ||
Infliximab detecting kit: Lot#B007, Diluent: human serum 1.
Although the kit was supposed to determine samples with an INF concentration of 1.0 µg/mL or greater as positive (+), the results of Table 2 indicate that many samples with a concentration of 1.0 µg/mL were not determined as positive (+), and that the test results for 1.0 and 1.3 µg/mL samples varied depending on the analyst. The result field of the tested plates turned yellow at the time of determination because of the color of the pooled human serum used for sample dilution, which may have affected the determination by visual inspection. Therefore, to evaluate the effect of human serum as a matrix on the visual inspection results, samples of innovator INF diluted with two different types of pooled human serum (human serum 1 and human serum 2) into five concentrations (0.3, 0.7, 1.0, 1.3 and 2.0 µg/mL) were tested using the kits (Lot#R905). Innovator samples diluted with 1% BSA in PBS were also tested as a reference. The test results obtained using the three types of matrices (i.e., human serum 1, human serum 2, and 1% BSA in PBS) are shown in Table 4 and Fig. 1. On the kit plates tested with samples diluted using 1% BSA in PBS, visual recognition of colored test sample lines was easier owing to the low intensity of background coloration, and more cases were observed wherein samples with a concentration of 1.0 and 1.3 µg/mL were determined as positive (+). When comparing two different types of pooled human serum, it was difficult to visually recognize the color of the samples diluted with human serum 1, which had a deep yellow color, and there were fewer cases where 1.0 µg/mL and 1.3 µg/mL samples were determined as positive (+). These results suggest that the color intensity of patient serum may affect the test results, as the color intensity of the sample line is visually compared to that of the cut-off control line in the kit.
The stock solutions of the innovator product were diluted with two different pooled sera, human serum 1 and human serum 2, or with 1% BSA in PBS to concentrations of 1.0, 1.3, and 2.0 µg/mL, and tested using the diagnostic kit. In each result field, the left line is the control line that indicates whether the test result is valid or invalid; the right line is the test line. In each result field, the upper images indicate the result of the cut-off control corresponding to 1.0 µg/mL, and the lower images indicate the results of test samples. The test line of the test sample was compared with that of the cut-off control based on the color intensity.
| Diluent | Analyst | Infliximab concentration (µg/mL) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.3 | 0.7 | 1.0 | 1.3 | 2.0 | ||||||||||||
| 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | ||
| Human Serum 1 | A | — | — | — | — | — | — | — | — | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | — | + | + | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | + | + | + | + | + | + | + | |
| Human Serum 2 | A | — | — | — | — | — | — | + | + | + | + | — | + | + | + | + |
| B | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | + | — | + | + | — | + | + | + | + | |
| 1% BSA in PBS | A | — | — | — | — | — | — | + | + | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | + | — | + | + | + | + | + | + | |
Infliximab detecting kit: Lot#R905.
In the three tests (Tables 2–4), the results for the innovator product were different, possibly because the lot of kits used in each test was not the same. Table 5 summarizes the test results for the innovator product in three lots. Figure 2 shows representative images of kit plates tested with innovator samples at a concentration of 1.0 µg/mL. In all three lots, there were cases in which the determination of 1.0 µg/mL samples differed depending on the analyst (Table 5). In particular, lot#B804 tended to yield negative results at a concentration of 1.0 µg/mL compared to that of the other two lots. There was also a slight difference in the clarity of the cut-off lines and test lines among different lots. These results indicate a possible lot-to-lot variation in performance along with serum characteristics. Notably, in actual clinical use, judgement around the threshold may vary depending on the background serum color, kit lot, and the analyst in charge of the visual inspection, because the test is usually performed using patient serum in one plate and is judged by one analyst.
| Lot No. of Infliximab detecting kit | Analyst | Infliximab concentration (µg/mL) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.3 | 0.7 | 1.0 | 1.3 | 2.0 | ||||||||||||
| 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | ||
| B804 | A | — | — | — | — | — | — | — | — | — | + | + | — | + | + | + |
| B | — | — | — | — | — | — | + | + | — | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | — | + | + | + | + | + | + | |
| R905 | A | — | — | — | — | — | — | — | — | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | — | + | + | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | — | + | + | + | + | + | + | + | |
| B007 | A | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + |
| B | — | — | — | — | — | — | + | — | + | + | + | + | + | + | + | |
| C | — | — | — | — | — | — | — | + | + | + | + | + | + | + | + | |
Diluent: human serum 1.
The stock solutions of the innovator product were diluted with human serum 1 to a concentration of 1.0 µg/mL, and tested using the diagnostic kit.
To support the notion that reactivity to the INF detection kit is not significantly different between the innovator and five BS products as suggested in Table 2, we compared the differences in the immunochemical properties of the innovator and BS products with other immunochemical methods; further, the reactivity of the innovator and five BS products on three kinds of ELISA kit for measuring serum INF concentrations was compared (Fig. 3). Among the three ELISA kits, one used immobilized recombinant human TNFα on the plate. The other two kits employed immobilized anti-INF monoclonal antibody on the plate. The WHO International Standard of INF was also used as a test sample. No differences were observed in the reactivity of the innovator, the five BS products, and the WHO standards among all ELISA kits. Thus, it is thought that the binding reactivities of INF with the anti-INF monoclonal antibody used in the kits, as well as recombinant TNFα, do not differ significantly among the innovators and five BS products. Neveu et al. also reported that the same quantitative results were obtained for the innovator and two BS products using four different ELISA kits.11) Similarly, it was reported that there were no significant differences in the binding affinity to TNFα between the innovator and two BS products.12) Although it is recognized that the reactivity of therapeutic protein products with the binding reagents used in immunochemical methods can vary depending on the manufacturing process, the BS products evaluated here were found to have immunochemical properties highly similar to those of the innovator product in terms of reactivity in the ELISA.
The stock solutions of the innovator, BS products, and WHO international standard were diluted with human serum at the indicated concentrations, and measured using each ELISA kit. The absorbance data show the mean value of duplicate wells for each concentration.
Inagaki et al. investigated the differences between the reactivities of the innovator and BS1[NK] with the diagnostic kit.10) Although they found cases in which samples with a concentration of 1.2 µg/mL were determined as negative (−), the reactivity of BS1[NK] with the diagnostic kit was almost the same as that of the innovator. Additionally, they indicated that there was variation among analysts as the color intensity of lines in the result field was low. They discussed that the kits should be designed such that the coloration of the test line is clearly visible in the event of a positive result. These results are in line with those of the present study. Regarding the color intensity of the lines, our results indicated that the color of pooled human serum used for dilution may affect the results. When testing actual serum samples collected from patients, the possibility of the color of the serum to be tested itself affecting the detection cannot be denied, suggesting the importance of proper patient serum preparation. Our results also indicated a possible difference in the clarity of the lines likely depending on the lot of the kit, which could be improved by the manufacturer.
The performance of the diagnostic kit was defined in the package insert based on the responses of the quality control standard. For sensitivity, when the control substance is tested at 0.7 µg/mL, it should be determined as negative (−), and when the control substance is tested at 1.3 µg/mL, it should be determined as positive (+). The performance on repeatability was also defined as that when the control substance was tested at 0.3 µg/mL three times simultaneously, all tested plates were determined as negative (−), and when testing the control substance at 2 µg/mL, three times simultaneously, all tested plates were determined as positive (+). Our test results on the innovator INF and the five BS products generally met these criteria even in the results of BS3[Pfizer] in Table 2, and thus mostly reproduced the manufacturer’s description, i.e., most of 0.7 and 1.3 µg/mL of both innovator and BS products were judged negative and positive, respectively.
The package insert of the kit also provided data regarding the evaluation of correlation between the ELISA and the diagnostic kit using serum collected from patients (156 cases) treated with the innovator (Remicade®). The concordance rate was reported to be 100% for samples with ELISA results less than 1.0 µg/mL, and 96.4% for samples with 1.0 µg/mL or higher. The results of ELISA and the diagnostic kit were discrepant in five cases with ELISA results of 1.09 to 1.37 µg/mL. This is in line with our test results, which showed that some test plates with samples at concentrations of 1.0 and 1.3 µg/mL were determined as negative (−) regardless of the type of INF product. In other words, a sample with a concentration of 1 µg/mL or higher may be diagnosed as negative (less than 1 µg/mL) in some cases. Because the intended use of this kit is qualitative evaluation, the bias up to ±30% may not be a major problem, considering that ±25% bias is accepted at the lower limit of quantification in quantitative analysis of serum concentration level in clinical studies by the ligand-binding assays according to the regulatory guideline for bioanalysis.13) Therefore, it is important to recognize that there is uncertainty (up to ±30% of bias) in the results obtained by the kit when using the results for judging the necessity of dose escalation.
As mentioned above, we found that the results of this kit around 1 µg/mL were variable depending on the lot of the kit, background serum color and analyst. The variability of each lot of the kit may be because of several factors such as quality of immobilized binding reagent, other binding reagent for color development, and pads for sample, conjugation, test line and absorbent. This can be improved by more tight control of the manufacturing process of the kit including materials used. The variation of serum color would be reduced by the cautious sampling of the blood. The variation among analysts for visual inspection can be reduced by training the analysts for consistent judgement or fix the analyst as an expert for judging the results.
In summary, the results of this study indicated that there was no significant difference in reactivity between the INF innovator and the five BS products in immunochromatography or ELISA in this study. Therefore, even in patients treated with BS products, it is possible to determine serum INF levels using a diagnostic kit developed for the innovator product. However, based on our results, treatment plans should be altered with caution, considering that the test results may vary with test conditions, including serum color, the lot of the kit, and analysts; further, INF levels of 1.0 µg/mL or slightly higher may be determined as negative (−) with the kit.
This research was supported by research Grants from the Japan Agency for Medical Research Development (AMED) under Grant Nos. JP19mk0101152 and JP22mk0101238.
The authors declare no conflict of interest.
