Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Clinical Chemistry and Laboratory Medicine (CCLM)

Published in Association with the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM)

Editor-in-Chief: Plebani, Mario

Ed. by Gillery, Philippe / Greaves, Ronda / Lackner, Karl J. / Lippi, Giuseppe / Melichar, Bohuslav / Payne, Deborah A. / Schlattmann, Peter


IMPACT FACTOR 2018: 3.638

CiteScore 2018: 2.44

SCImago Journal Rank (SJR) 2018: 1.191
Source Normalized Impact per Paper (SNIP) 2018: 1.205

Online
ISSN
1437-4331
See all formats and pricing
More options …
Volume 55, Issue 8

Issues

Preparation, calibration and evaluation of the First International Standard for human C-peptide

Melanie Moore / Thomas Dougall / Jackie Ferguson / Peter Rigsby / Chris Burns
Published Online: 2017-05-24 | DOI: https://doi.org/10.1515/cclm-2017-0029

Abstract

Background:

Measurement of C-peptide by immunoassay contributes to the diagnosis of a number of disorders related to β cell function. Stocks of the current international reference reagent (IRR) for C-peptide, used to calibrate these immunoassays, are exhausted, and this report summarises the international study to establish a replacement World Health Organization (WHO) international standard (IS) to maintain the availability of a globally available reference material and support efforts to standardise C-peptide assays.

Methods:

The study was conducted in three phases; phase I involved the assignment of a value to a primary calibrant in mass units by amino acid analysis and phase II applied this value to the calibration of a candidate standard, 13/146, by reversed phase high-performance liquid chromatography (RP-HPLC) assay. In phase III, the candidate standard was compared to the first IRR by current immunoassays to assess its suitability to serve as an IS.

Results:

Calibration of the candidate standard by RP-HPLC gave a final estimated content of 8.64 μg/ampoule with expanded uncertainty of 8.21–9.07 μg/ampoule (95% confidence; k=2.45). The candidate standard also appears sufficiently stable to serve as an IS, based on HPLC analysis of accelerated thermal degradation samples of 13/146, and was also shown to have appropriate immunological activity. A difference in bias approach was used to assess the commutability of 13/146 with human serum and urine samples. With the exception of two laboratories, the candidate standard demonstrated commutability with respect to the serum and urine samples included in this study.

Conclusions:

The candidate standard, 13/146, is suitable to serve as the First International Standard for human C-peptide, and it has been formally adopted by the Expert Committee on Biological Standardisation of the WHO.

This article offers supplementary material which is provided at the end of the article.

Keywords: commutability; human C-peptide; immunoassay; WHO international standard

Introduction

C-peptide is synthesised in pancreatic β cells as the connecting peptide for the A and B chains of insulin in the proinsulin molecule. Proinsulin is converted to insulin and C-peptide by a process of enzymatic cleavage, and these are then secreted into the circulation by the β cells. Insulin and C-peptide are secreted in equimolar amounts, with C-peptide having a longer half-life in the peripheral circulation than insulin [1]. Therefore, measurements of C-peptide can provide a reliable indication of pancreatic insulin secretion, even in the presence of exogenous insulin treatment. Human C-peptide is measured by immunoassay in patient serum and plasma. It can also be measured in urine as a ratio of urinary C-peptide: urinary creatine, providing a less invasive method for sample collection and patient monitoring. These measurements have a number of important clinical uses where the monitoring of β cell function and endogenous insulin levels is required (reviewed in [1]). For example, C-peptide measurements play a key role in the differentiation between type I and type II diabetes and in the monitoring of residual β cell function over time [2]. C-peptide measurements can also be used to aid in the diagnosis of hypoglycaemia and insulinoma, in monitoring diabetes treatments [3], and also have the potential to serve as a marker of residual pancreatic tissue after pancreatectomy, or to verify effectiveness of pancreatic islet cell transplantation [4].

These types of diagnoses and monitoring of treatment require the use of clinical practice guidelines, and these should be based on evidence derived from data obtained in multiple studies often using a range of different C-peptide immunoassays. For these data to be combined to derive guidelines, the assays should provide the same results for a given patient sample (i.e. be harmonised). However, although these assays have been traceable to a World Health Organization (WHO) international reference reagent (IRR) for human C-peptide (National Institute for Biological Standards and Control [NIBSC] code 84/510), which has been available for almost 30 years for the purposes of calibrating C-peptide immunoassays [5], there remains significant between-assay method variability for measurements in patient samples. Stocks of this standard are now exhausted, and there is a requirement for a replacement to maintain the provision of a reference material for C-peptide and to support important initiatives to improve the between method variability of C-peptide immunoassays. In addition, the replacement of the IRR provides an opportunity to calibrate a new international standard (IS) with current physicochemical methods allowing for the assignment of a more accurate mass content than was possible 30 years ago, when the IRR was established. The proposal to develop the First International Standard for human C-peptide was endorsed by the WHO Expert Committee on Standardisation (ECBS) at its meeting held in November 2010. This report summarises the preparation, calibration and evaluation of a candidate preparation and its formal adoption and establishment by WHO as the First International Standard for human C-peptide. Full details of the WHO report are available online (a link to reference document BS2256 is available at http://www.who.int/iris/handle/10665/197773).

Materials and methods

Candidate IS

Bulk material:

Synthetic human C-peptide, 75 mg, was generated by custom peptide synthesis (Almac, Elvingston Science Centre, UK) according to the C-peptide amino acid sequence, EAED LQVG QVELG GGPG AGSL QPLA LEGS LQ (NCBI ref P01308).

Formulated materials:

C-peptide was formulated in a prefiltered buffer containing 10 mM sodium phosphate pH 7.0 and 0.5% (w/v) trehalose and dispensed into 3 mL ampoules as 0.5 mL aliquots, freeze-dried and sealed under nitrogen at NIBSC, according to procedures described by the WHO for the preparation of International Biological Standards [6]. A limited number of vials of primary calibrant, containing a nominal 250 μg per ampoule and coded primary calibrant (PC01), were also prepared. Briefly, C-peptide was formulated in a prefiltered buffer containing 10 mM sodium phosphate pH 7.0, 0.5% (w/v) trehalose and dispensed into 3 mL crimp vials as 0.5 mL aliquots, freeze-dried and sealed under nitrogen.

Collaborative study design

The candidate standard was estimated to contain approximately 8 μg of formulated synthetic C-peptide. Because this cannot easily be accurately measured directly by physicochemical methods, the collaborative study was devised to comprise three phases. Phase I involved establishing a PC01, consisting of a limited number of vials containing approximately 250 μg C-peptide, which was assigned a defined value using amino acid analysis (AAA). Phase II involved calibration of the candidate standard 13/146 in terms of PC01 by high-performance liquid chromatography (HPLC). In this phase of the study, the effect of accelerated thermal degradation (ATD) on the candidate standard was also assessed by HPLC to enable a prediction of its long-term stability. The purposes of phase III were to provide confirmatory data by immunoassay and to assess the suitability of the candidate standard to serve as an IS by comparison with the current standard, 84/510, and existing local standards. In addition, human serum and urine samples were incorporated into phase III as part of a small commutability study in order to assess the impact of the introduction of the candidate standard (formulated synthetic C-peptide) on the routine measurement of C-peptide in native samples. Participants from 24 laboratories in 10 countries took part in the study (full details are provided in the WHO report).

Collaborative study materials:

Participants were provided with vials of PC01, 13/146, ATD samples of 13/146, the first IRR (84/510) and serum/urine samples depending on the phase of study they participated in (details are provided in Supplementary Table A1). In order to provide a range of C-peptide values, both fasting (overnight) and nonfasting serum and urine samples were obtained for inclusion in phase IIIb. Samples were coded, and a total of 16 human serum samples and 16 human urine samples were made available to participants, including four C-peptide negative serum and four C-peptide negative urine samples spiked with 13/146 to provide (or provide after dilution of urine samples) 5 ng/mL, 2.5 ng/mL, 1.25 ng/mL and 0.625 ng/mL C-peptide. Serum samples were kindly collected by Dr G. Wark (UK NEQAS, Surrey Pathology Services, UK), purchased from FirstLink (UK) Ltd or obtained from NIBSC in-house volunteer blood donors. Serum from UK NEQAS and FirstLink (UK) Ltd was frozen and sent to NIBSC on dry ice, where it was thawed, dispensed into 0.5 mL or 1 mL aliquots and stored at −80 °C. Blood from in-house NIBSC blood donors was allowed to clot at room temperature for 1 h, before transferring to 4 °C overnight. The samples were then centrifuged at 2000×g for 15 min at 4 °C, the serum removed and then dispensed into 0.5 mL aliquots and stored at −80 °C. Urine samples were purchased from Sera Laboratories International Ltd (West Sussex, UK) and obtained from NIBSC in-house volunteer donors. Urine samples from Sera Laboratories were frozen and sent to NIBSC on dry ice. As urine samples are extensively diluted (1:5–1:20) before use in an immunoassay, after thawing, samples were dispensed into 50 μL aliquots to reduce the number of preassay pipetting steps required by a laboratory. These were then stored at −80 °C. Urine samples from in-house NIBSC donors were dispensed into 50 μL aliquots on the day of donation and stored at −80 °C. The inclusion of serum and urine samples in this study was approved by a local ethics committee. In order to maintain sample stability, both serum and urine samples were shipped to participants on dry ice. Participants were requested to store samples at −80 °C prior to use in their assay.

Methods

Details of analytical methods are summarised in Supplementary Table A2. Full assay details and raw data were reported but are not reproduced here.

Phase I assays:

Participants were requested to derive estimates of the human C-peptide content of the primary calibrant, PC01, by AAA using their in-house method. Three vials of the primary calibrant were provided, and participants were requested to carry out a minimum of two analytical runs for each vial.

Phase II assays:

Using the assigned value of 209 μg per ampoule for PC01, derived from data provided in phase I, participants in phase II were requested to provide, in triplicate, estimates of the C-peptide content of the candidate standard 13/146 and its ATD samples by comparison with PC01 using HPLC. An outline HPLC protocol was provided for guidance. Briefly, the protocol asked participants to perform reversed phase high-performance liquid chromatography (RP-HPLC) using the following equipment (or equivalent): a C8 or C18 HPLC column, pore size 150–300 A, particle size 5–7 μm, with a column oven to maintain constant column temperature if available; a chilled auto-sampler to maintain samples at 4 °C if available; a binary pump with mobile phase A of 0.1 M sodium phosphate monobasic dihydrate and 0.5% (v/v) phosphoric acid and mobile phase B of acetonitrile with a flow rate of 1 mL/min. Participants were recommended to dilute primary calibrant in HPLC-grade water to a concentration of 40 μg/mL and inject volumes of 100, 75, 50, 25 and 12.5 μL to provide a standard curve with 4, 3, 2, 1 and 0.5 μg C-peptide primary calibrant, respectively. Participants were recommended to reconstitute 13/146 ampoule contents in 0.5 mL HPLC-grade or double-distilled water and inject 100 μL in triplicate, using a linear gradient of 5%–50% mobile phase B in mobile phase A to elute C-peptide between 29 and 31 min. Absorbance was monitored at 214 nm. C-peptide content of each ampouled preparation was calculated from the standard curve of the primary calibrant by linear regression analysis.

Phase III assays:

Participants were requested to carry out the assay normally in use in their laboratory and, where possible, to perform two independent assays using fresh ampoules, including all the preparations allocated at no less than five dose levels in the linear part of the dose-response curve. Specific handling instructions for the materials were provided to participants describing a core range of dose levels that were to be included in the assay by all laboratories.

Statistical analysis:

Participants’ own estimates of C-peptide content were collated from phases I, II and IIIa and b for analysis at NIBSC. The relative contents of the ATD samples determined by immunoassay in phase IIIa were analysed with a parallel line model using EDQM CombiStats Software Version 5.0 [7]. These results and relative contents by HPLC from phase II were used to fit an Arrhenius equation relating degradation rate to absolute temperature, assuming first-order decay, and hence predict the degradation rates when stored at −20 °C [8].

For the assessment of commutability in phase IIIb, reported concentrations (ng/mL) for all samples were log10 transformed in order to achieve approximately constant bias over the range of concentrations used. For each laboratory pair, the difference between the C-peptide concentrations reported by each laboratory for each serum or urine sample was calculated to provide the bias value for that sample. The mean and standard deviation of the bias values, for serum or urine samples, were determined, and the standard deviations pooled to set commutability criteria as ±2 SDserum or ±2 SDurine for the acceptable difference in bias between reference materials and serum or urine samples, respectively. The reference material, at each common dilution, was considered commutable for the laboratory pair if the difference between the bias of the reference material determinations, and the mean bias of the serum (or urine) samples was within the relevant defined limits.

Between-laboratory variability for individual samples in phase IIIb has been assessed using geometric coefficients of variation (GCV={10s−1}×100%, where s is the standard deviation of the log10 transformed estimates).

Results

Amino acid analysis (AAA)

The estimated content of the primary calibrant, PC01, was reported by six laboratories in phase I of the study. With the exception of laboratory 5 who received one vial of PC01, each laboratory received three vials of PC01 and performed duplicate or triplicate analysis, providing a total of 36 AAAs of PC01. The mean estimated C-peptide content per vial of PC01 (μg/vial) is shown in Table 1. The results from each laboratory were in good agreement, with mean laboratory estimates of PC01 ranging from 194 to 215 μg/vial. The mean content from each laboratory was combined to give a final estimated content of 209 μg/vial (95% confidence limits: 199–219 μg/vial).

Table 1:

Mean estimated C-peptide content of PC01 (μg/vial) by AAA.

High-performance liquid chromatography (HPLC)

Estimated content of candidate standard 13/146 in terms of PC01

The estimated C-peptide content of the candidate standard, 13/146, was reported by seven laboratories. In total, 57 HPLC assays were performed for the calibration of 13/146, and data are summarised in Table 2. Data used for analysis were that reported by participants using the peak area of the main C-peptide peak to calibrate the candidate standard in terms of PC01 (an example chromatogram is included in Supplementary Figure A1). Some participants (laboratories 8, 10 and 11) noted a “shoulder” peak to the main C-peptide peak. This was included in calculations by laboratory 8 but excluded from content calculations by laboratory 10. Laboratory 11 provided data both excluding and including the shoulder peak and showed no difference in final ampoule content estimates. Results were in good agreement between laboratories, with laboratory mean estimates of 13/146 ranging from 8.45 to 8.89 μg/ampoule, providing a final estimate (mean of laboratory means) of 8.64 μg/ampoule. The standard error of the value assigned to PC01 (1.88%) and the homogeneity of filling weight (0.23% CV) were combined with the HPLC standard error (0.66%) to give a combined standard uncertainty of 2.01% and an uncertainty estimate of 0.174. This gave a final estimate for 13/146 of 8.64 μg/ampoule with expanded uncertainty of 8.21–9.07 μg/ampoule (95% confidence; k=2.45).

Table 2:

Mean estimated C-peptide content of candidate standard 13/146 by HPLC relative to PC01.

Stability of 13/146 based on thermally accelerated degradation study by HPLC

In addition to providing estimates of C-peptide content of the candidate standard, six laboratories also estimated C-peptide content of ATD samples of 13/146 that had been stored at elevated temperatures for a period of 7 months. Estimated C-peptide content of ATD samples was reported using the peak area of the main C-peptide peak. Three laboratories (8, 9 and 10) noted an additional minor peak eluting after the main C-peptide peak in ATD samples, and that this peak was increased significantly in sample P (+45 °C ATD sample). It is possible that the N-terminal glutamine residue of the peptide may have formed a pyroglutamate upon thermal degradation or long-term storage, which may have given rise to a peak with increased retention time. These impurity peaks were excluded from the content calculations of ATD samples.

In order to assess the stability of the candidate standard, estimates were made of the relative C-peptide content of the ATD samples stored at elevated temperature for 7 months in comparison to the combined estimate for ampoules that had been stored continuously at −20 °C. These results are summarised in Table 3. The loss of C-peptide (main peak response) after storage at elevated temperatures can be used to predict the stability of WHO IS. The rate of loss of C-peptide is related to temperature using the Arrhenius equation, which can then be used to predict the long-term stability at lower temperatures, assuming first-order decay [8] (Table 3). A yearly loss of 0.07% per year at −20 °C storage conditions was predicted using this approach (equivalent to 0.006 μg/year) and indicates that the candidate standard is likely to be highly stable under long-term storage conditions at −20 °C and, as such, is sufficiently stable to serve as an IS.

Table 3:

Mean estimated C-peptide content (μg/ampoule) in thermally degraded samples of 13/146 assessed by HPLC.

Immunoassays

Estimates of C-peptide immunoreactivity in 84/510 and 13/146

Immunoassay data were contributed by 13 laboratories, five of which used more than one method. Where this was the case, the laboratory code has been subdivided for method differences, for example, 17a, 17b, 17c and 17d. Mean concentrations (ng/mL) of dilutions of the first IRR 84/510 and the candidate standard, 13/146, from both phases IIIa and IIIb were provided by participants, as calculated by comparison with kit standards using their in-house method. Estimated ampoule contents (μg/ampoule) have been calculated from these dilutions, and these are summarised in Table 4. Assay results by laboratory 24 g were excluded from further analysis as they represented anomalously low values for ampoules of the candidate standard that were a result of a suspected procedural fault. Laboratory 19 compared reconstitution of 13/146 in either PBS/BSA 0.1% (19a) as suggested in the study protocol or horse serum (19b) as suggested by the kit manufacturer. Results from laboratory 19a (buffer) were markedly higher for 84/510 and 13/146 compared with the manufacturer’s recommended matrix, and these were therefore removed from the overall analysis of laboratory estimates as they were deemed anomalous for that assay system.

Table 4:

Mean estimated C-peptide content (μg/ampoule) of 84/510 and 13/146.

Laboratory estimates for ampoule content were in reasonable agreement, with a few laboratories reporting notably higher ampoule content for 13/146 (laboratories 18b and 21–24a), but were not provided with ampoules of 84/510 for comparison due to limited stocks. The mean of laboratory estimates of 84/510 ampoule content gave 9.20 μg/ampoule (95% confidence limits 8.77–9.63), approximately 8% lower than the assigned content of 10 μg/ampoule for 84/510. The mean of laboratory estimates of 13/146 ampoule content gave 9.78 μg/ampoule (95% confidence limits 9.31–10.25), which is approximately 13% higher than the estimated content of 8.64 μg/ampoule for 13/146, as assigned by HPLC in phase II of this study. It should be noted that the C-peptide content of 84/510 was assigned by immunoassay estimates in a limited study (five laboratories with a total of seven immunoassays) in the late 1980s [5]. The assigned content of 84/510 is 10 μg per ampoule, but it is very possible this may be an overestimation. Taking the assigned content of 13/146 from phase II (8.64 μg/ampoule) and the ratio of immunoreactivity between 84/510 and 13/146 observed in this study, which is 0.94, it is likely that the content of 84/510 may be closer to 8.1 μg/ampoule. Because the manufacturer’s kit standards in this study are calibrated in terms of 84/510, with its assigned value of 10 μg/ampoule, it is perhaps not surprising that immunoassay estimates of the content of the candidate standard are higher than would be expected on the basis of the physicochemical assays. Manufacturers should therefore be made aware of the potential impact that replacement of 84/510 with 13/146, which has been assigned a more accurate value using physicochemical methods, may have on their assay system.

Stability of 13/146 assessed by immunoassay

Five laboratories tested 13/146 and ATD samples of 13/146 that were stored at elevated temperatures of +4, +20, +37 and +45 °C for 18 months. Using 13/146 stored continuously at −20 °C as a baseline, estimates of the relative ampoule contents of 13/146 ATD samples were calculated (data not shown). Laboratories were in good agreement, giving a mean percentage C-peptide content of 100% (95% confidence limits of 99–101) for +4 °C, 100% (98–103) for +20 °C, 102% (99–102) for +37 °C and 97% (91–102) for +45 °C. Compared with HPLC analysis of degradation samples, there appears to be very little effect of thermal degradation on the immunoreactivity of the candidate standard, with no loss in immunoreactivity detected at +4, +20 or +37 °C elevated temperatures, and a very small percentage loss of 3% at +45 °C. This limited loss in activity did not allow a predicted yearly loss of activity to be determined from immunoassay data but did support the conclusion that 13/146 is sufficiently stable to serve as an IS.

Commutability of 13/146 and 84/510

Immunoassay data from phase IIIb of the study, in which serum and/or urine samples were included in assays alongside ampouled preparations, were reported by 14 laboratories (serum samples) and 11 laboratories (urine samples). Three serum samples and two urine samples were obtained from type I diabetics for inclusion in the study. As expected, values reported for these samples were described as zero or at/below the limit of detection for all immunoassays. These samples were then removed from further analysis of commutability. The statistical analyses used in the assessment of commutability of the ampouled preparations are described in the section “Statistical analysis”. Similar to the analysis in phase IIIa, for laboratory 19, only ampoule estimates from laboratory 19b results were further analysed as these represented ampouled preparations reconstituted in assay buffer as recommended by the manufacturer (horse serum matrix). In addition, assay run 2 results for samples AC–AF (13/146 spiked into C-peptide negative urine) performed by laboratory 17 in assays 17a, 17b, 17c and 17d were also excluded from further analysis due to potential technical error in the preparation or dilution of the samples.

Average bias in serum and urine sample results

Measured concentrations (ng/mL) for serum samples, urine samples and reference material dilutions are provided in Supplementary Tables A3–A7. The measured concentrations for all serum and urine samples were log10 transformed, and for each laboratory pair, the mean and standard deviation of the bias were determined. The mean bias for serum and urine samples results for each laboratory pair is shown in Table 5. The magnitude of the average bias ranged from zero (no difference between laboratories) to 0.230 (serum samples; laboratory pair 18a and 20) and 0.136 (urine samples; laboratory pair 16 and 18b). Standard deviations in bias calculated for each laboratory pair were pooled and used to set commutability criteria as ±2 SDserum or ±2 SDurine, to give the acceptable difference in bias between reference materials and serum or urine samples. These calculations gave acceptable ranges of ±0.124 and ±0.200 for serum and urine samples, respectively.

Table 5:

Average bias in log10 serum and urine sample results for each laboratory pair.

Commutability of 13/146 and 84/510

Tables 6 and 7 summarise the proportion of reference material dilutions concluded to be noncommutable for each laboratory pair (see Supplementary Tables A5 and A6 for details of these reference material dilutions). For example, where there are four dilutions of 13/146, a value of 0.50 indicated that two of the dilutions were noncommutable (coloured green) in that laboratory pair. The tables clearly suggest that, with exception of laboratories 19 and 20, the first IRR 84/510 and the candidate First International Standard 13/146 demonstrate acceptable commutability with respect to serum and urine samples when commutability is assessed in this way. Combining the percentages of each reference material dilution concluded to be commutable (Supplementary Table A7) gives a mean percentage of 79.8% and 100% of 13/146 dilutions commutable with serum and urine samples, respectively, and 81.8% and 99.2% of 84/510 dilutions commutable with serum and urine samples, respectively. It should be noted that the analysis of commutability using bias, as described in this study, is only one of a number of methods that may be used to assess this parameter. Here, the commutability criteria have been assigned based on the statistical analysis of the bias, rather than any clinically derived criteria. It is possible that a commutability criteria derived in a different way may lead to an alternative conclusion with regard to the commutability of these reference materials.

Table 6:

Proportion of 13/146 dilutions concluded to be noncommutable for each laboratory pair.

Table 7:

Proportion of 84/510 dilutions concluded to be noncommutable for each laboratory pair.

Phase IIIb results were also used to compare whether reporting serum and urine results relative to kit standards, the first IRR 84/510 or the candidate First International Standard 13/146 altered the variability in results. The between-laboratory variability (the % GCV values) for each serum and urine sample was calculated from results as reported by participants and also from results expressed relative to the corresponding laboratory mean for 13/146 or 84/510 (Table 8). Similar values (pooled GCV ~20%) were obtained regardless of the reference (in-house calibrators, 13/146 or 84/510) used. A subset of these data, the between-laboratory variability of 13/146 spiked in C-peptide negative serum samples or urine samples (5–0.625 ng/mL), also showed no difference in % GCV whether calibrated to in-house standards, 13/146 or 84/510. These spiked samples were included to assess the performance of dilutions of standards in different matrices, and it can be concluded that the candidate standard behaves in a similar manner in both matrices included in the study. However, the pooled % GCV data for both serum and urine samples demonstrate an increase in variability between laboratories as the spike in C-peptide concentration is decreased from 5 ng/mL to 0.625 ng/mL. Overall, these data indicate, perhaps as expected given the purified and formulated nature of 13/146, that the introduction of the candidate standard, 13/146, for the calibration of immunoassays for C-peptide is unlikely to change the between-method variability currently observed for results for patient samples. In terms of absolute values, the impact of introducing 13/146 on the C-peptide content of patient samples will vary depending on laboratory. On average (excluding laboratory 20), absolute values for serum samples expressed relative to 13/146 were approximately 95% of those values reported by participants and 90% for urine samples (data not shown). As an example, serum A was reported as 4.9 ng/mL by laboratory 16, but was recalculated as 4.6 ng/mL when expressed relative to 13/146 using data provided by the laboratory. Although data indicate that introduction of the candidate standard 13/146 is unlikely to result in increased variability between laboratories, individual manufacturers may need to make small adjustments to their local standards.

Table 8:

Summary of between-laboratory variability (% GCV) for serum and urine samples.

Discussion and conclusions

Measurement of C-peptide by immunoassay contributes to the diagnosis of a number of disorders related to β cell function. Stocks of the current IRR for C-peptide, used to calibrate these immunoassays, are exhausted, and this report summarises the international collaborative study to establish a replacement WHO IS to maintain the availability of a globally available reference material to support efforts to standardise C-peptide assays. Thus, it has been recognised for some time that C-peptide results from different immunoassays do not always agree and that there is a need for harmonisation of these assays. There have been a number of recent reports demonstrating the potential role of calibrators, prepared in a sample matrix and value-assigned using a reference method, in improving the between-method agreement of C-peptide measurements in patient samples (9–12). This effort is now underway, with significant progress reported by the NIDDK C-peptide standardisation committee (R. Little, personal communication). We fully support these initiatives and believe they will be essential for the harmonisation of C-peptide assays. Indeed, the establishment of the WHO IS, described in this report, is unlikely to improve the current between-method agreement for C-peptide assays, and this is perhaps not unexpected given the nature of the material. However, despite this fact, it was decided by the Expert Committee on Biological Standardisation (ECBS) of WHO, at their 2010 meeting, that having made a reference material available to manufacturers for nearly 30 years and with no certainty that an alternative material would be available in the near future, there was an ongoing responsibility to this community to maintain the availability of such a reference material for the continuing development, calibration and monitoring of C-peptide assays. It was, however, also recognised that the anticipated uses of this material may change once the harmonisation and/or standardisation efforts described above have been successfully implemented.

The study was conducted in three phases; phase I involved the assignment of a value to a primary calibrant in mass units by AAA, and phase II applied this value to the calibration of a candidate standard by RP-HPLC assay. In phase III, the candidate standard was compared to the First IRR by current immunoassays to assess its suitability to serve as an IS. In this phase, participants were also requested to determine the C-peptide concentration of 16 serum and 16 urine samples in order to assess the impact of the introduction of the candidate preparation on the routine measurement of C-peptide in native samples.

Laboratories were in good agreement in phases I and II, and calibration of the candidate standard by comparison with PC01 in RP-HPLC gave a final estimate of the content of the candidate standard of 8.64 μg/ampoule with expanded uncertainty of 8.21–9.07 μg/ampoule (95% confidence; k=2.45). The candidate standard 13/146 also appears to be sufficiently stable to serve as an IS, based on HPLC analysis of ATD samples of 13/146. The candidate standard has also been shown to have appropriate immunological activity, and there was reasonable agreement between laboratories for the mean estimates of the immunoreactivity of 13/146 and also for the First IRR 84/510. Compared to the assigned value for 13/146 of 8.64 μg/ampoule, estimates of the content of 13/146 were higher when measured using immunoassays. These assays are calibrated in terms of 84/510, which was assigned by immunoassay estimates in a limited study (five laboratories with a total of seven immunoassays) in the late 1980s, and it now seems likely that its assigned content was an overestimation. As a result, the adoption of 13/146 as a standard for the calibration of immunoassays (the IRR 84/510 is now unavailable) would necessitate a small adjustment of local standards. Demonstration of commutability with patient samples is an important attribute for a diagnostic reference material. We have used elements of the difference in bias approach recently reported by Korzun et al. [9] to assess the commutability of 13/146 with serum and urine samples. With the exception of two laboratories (laboratory 19 and 20), both the First IRR 84/510 and the candidate standard 13/146 demonstrated commutability with respect to the serum and urine samples included in this study.

In conclusion, there are ongoing efforts to harmonise C-peptide immunoassays, including the development of isotope-dilution liquid chromatography-mass spectrometry (LCMS) reference methods [10, 11] and a certified reference material [12] for the calibration of these methods, in addition to investigations into the use of panels of serum calibrators for standardisation/harmonisation [13, 14]. It is anticipated that the First International Standard will complement these efforts to aid in the standardisation and harmonisation of C-peptide immunoassays. ISs are developed to be highly stable and widely available and are expected to last between 10 and 15 years. This will enable a constant supply of stable reference material, obviating against potential discontinuity of unitage that is inherent in any replacement exercises for serum calibrators. The First International Standard for C-peptide comprises formulated synthetic C-peptide material that has been shown in this study to demonstrate both stability and commutability and therefore could be a valuable resource to reduce “drift” in these exercises, in addition to its current use as a standard for the calibration of C-peptide immunoassays.

The ECBS of the WHO formally adopted the preparation 13/146 as the First International Standard for human C-peptide. The material is now available on request from NIBSC.

Acknowledgments

We gratefully acknowledge the important contributions of all the participants, the volunteer donors who donated blood and urine for inclusion in this study, Mr Hirofumi Koide of the Japanese Committee for Clinical Laboratory Standards for organising participating laboratories from Japan and finally the Centre for Biological Reference Materials, NIBSC, for preparation of the ampouled materials.

References

  • 1.

    Jones AG, Hattersley AT. The clinical utility of C-peptide measurement in the care of patients with diabetes. Diabet Med 2013;30:803–17. PubMedCrossrefGoogle Scholar

  • 2.

    Oram RA, Jones AG, Besser RJ, Knight BA, Shields BM, Brown RJ, et al. The majority of patients with long-duration type 1 diabetes are insulin microsecretors and have functioning beta cells. Diabetologia 2014;57:187–91. Web of ScienceCrossrefPubMedGoogle Scholar

  • 3.

    Cryer PE, Axelrod L, Grossman AB, Heller SR, Montori VM, Seaquist ER, et al. Evaluation and management of adult hypoglycemic disorders: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 2009;94:709–28. PubMedWeb of ScienceCrossrefGoogle Scholar

  • 4.

    Ryan EA, Paty BW, Senior PA, Lakey JR, Bigam D, Shapiro AM. Beta-score: an assessment of beta-cell function after islet transplantation. Diabetes Care 2005;28:343–7. CrossrefPubMedGoogle Scholar

  • 5.

    Bristow AF, Gaines Das RE. WHO International reference reagents for human proinsulin and human insulin C-peptide. J Biol Standardisation 1986;16:179–86. Google Scholar

  • 6.

    WHO Tech Rep Ser No 800. 1990;181–214. 

  • 7.

    CombiStats v5.0, EDQM – Council of Europe. Available at www.combistats.eu

  • 8.

    Kirkwood TL. Predicting the stability of biological standards and products. Biometrics 1977;33:736–42. PubMedCrossrefGoogle Scholar

  • 9.

    Korzun WJ, Nilsson G, Bachmann LM, Myers GL, Sakurabayashi I, Nakajima K, et al. Difference in bias approach for commutability assessment: application to frozen pools of human serum measured by 8 direct methods for HDL and LDL cholesterol. Clin Chem 2015;61:1107–13. Web of ScienceCrossrefPubMedGoogle Scholar

  • 10.

    Rogatsky E, Balent B, Goswami G, Tomuta V, Jayatillake H, Cruikshank G, et al. Sensitive quantitative analysis of C-peptide in human plasma by 2-dimensional liquid chromatography-mass spectrometry isotope-dilution assay. Clin Chem 2006;52:872–9. PubMedCrossrefGoogle Scholar

  • 11.

    Rodriguez-Cabaleiro D, Stockl D, Kaufman JM, Fiers T, Thienpont LM. Feasibility of standardisation of serum C-peptide immunoassays with isotope-dilution liquid chromatography-tandem mass spectrometry. Clin Chem 2006;52:1193–6. CrossrefGoogle Scholar

  • 12.

    Kinumi T, Goto M, Eyama S, Kato M, Kasama T, Takatsu A. Development of SI-traceable C-peptide certified reference material NMIJ CRM 6901-a using isotope-dilution mass spectrometry-based amino acid analyses. Anal Bioanal Chem 2012;404:13–21. Web of ScienceCrossrefPubMedGoogle Scholar

  • 13.

    Weidmeyer HM, Polonsky KS, Myers GL, Little RR, Greenbaum CJ, Goldstein DE, et al. International comparison of C-peptide measurements. Clin Chem 2007;53:784–7. Web of ScienceCrossrefPubMedGoogle Scholar

  • 14.

    Little RR, Rohlfing CL, Tennill AL, Madsen RW, Polonsky KS, Myers GL, et al. Standardization of C-peptide measurements. Clin Chem 2008;54:1023–6. PubMedCrossrefWeb of ScienceGoogle Scholar

Supplemental Material:

The online version of this article (DOI: https://doi.org/10.1515/cclm-2017-0029) offers supplementary material, available to authorized users.

About the article

Corresponding author: Dr. Melanie Moore, National Institute for Biological Standards and Control, Blanche Lane, Potters Bar, Herts, EN6 3QG, UK


Received: 2017-01-13

Accepted: 2017-04-03

Published Online: 2017-05-24

Published in Print: 2017-07-26


Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

Research funding: None declared.

Employment or leadership: None declared.

Honorarium: None declared.

Competing interests: The funding organisation played no role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.


Citation Information: Clinical Chemistry and Laboratory Medicine (CCLM), Volume 55, Issue 8, Pages 1224–1233, ISSN (Online) 1437-4331, ISSN (Print) 1434-6621, DOI: https://doi.org/10.1515/cclm-2017-0029.

Export Citation

©2017 Melanie Moore et al., published by De Gruyter, Berlin/Boston. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. BY-NC-ND 3.0

Supplementary Article Materials

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

[1]
Sebastian Hörber, Peter Achenbach, Erwin Schleicher, and Andreas Peter
Biotechnology Advances, 2019

Comments (0)

Please log in or register to comment.
Log in