Birminghamquality.org.uk

Standardisation of Human Growth Hormone (hGH) Assays
Report of a meeting held at National Institute for Biological Standards and Control(NIBSC), South Mimms, Potters Bar, Hertfordshire, UK Manufacturers of hGH assay devices, clinicians, laboratory scientists, representatives from NIBSC, UK NEQASand BIVDA attended. Appendix A shows the list of participants.
Purpose and scope of the meeting
The background to the meeting was the poor agreement in hGH results using methods from differentmanufacturers, and the consequent risk of patient mismanagement. The specific points to be addressed were: • What calibrant should be used for hGH immunoassays?• What concentration units should be used?• What antibody specificities should be used?• What other steps, e.g. assay design to avoid interference from hGH binding proteins) should be taken to It was acknowledged that progress in these areas required planned and co-ordinated action between IVDmanufacturers, users of hGH assay devices, and standardisation and quality assurance agencies.
Recommendations from the meeting would have advisory rather than formal status, but would represent a broadand authoritative view that could be channeled to the WHO ECBS. It was also acknowledged that the meetingdid not represent all manufacturers in the global hGH assay market, although all but one manufacturer marketinghGH devices in the UK was included.
Presentations
What does the clinician require from an hGH assay?
Prof. M Preece ( Institute for Child Health, London) From the paediatric endocrinologists point of view, the reliability of GH assays was only one of severaldifficulties in using hGH assays in assessing short stature. These included (a) the practical inconvenience ofphysiological methods of assessing hGH secretion (e.g. day profiles, urinary excretion), (b) the non-physiological nature of commonly used provocative tests (e.g. insulin or clonidine stimulation), (c) the failure toallow for the effect of gonadal steroid hormone secretion on hGH response, (d) the non-reproducibility of hGHresponses, and (e) the lack of well-validated reference data. Improvements in hGH assay reliability wouldhowever provide a more firm basis for these problems to be addressed.
Current problems in hGH assays as viewed from UK NEQAS
Major advances in assay technology over the past 20 years had made no improvement in between laboratoryagreement, and results could show a difference of up to two fold. Incorrect calibration, and differences in assayspecificity were identified as contributory factors. Preliminary data showed that correct calibration of assaysusing rDNA-derived hGH in place of the pituitary-derived hGH calibrator currently used would improve inter-method agreement.
International standards for hGH –preparation and properties
The history of hGH International Standards was briefly reviewed. The first standard (66/217) and its more puresuccessor (80/505) were pituitary derived. Their content was assigned in terms of units of bioactivity, the Annual Review for 2000 - Appendix (Growth hormone) original numerical value of which was arbitrarily defined. The mass content of these preparations was neverdefined, although many users incorrectly use the approximate mass content (given only as a guide) as theassigned value. Since 1994 hGH standards (88/624 and its successor 98/574) have been prepared by r-DNAtechnology. These standards contain 22kDa hGH of >95% purity, with a defined specific activity of 3.0 IU/mg.
Content is defined in terms of mass, initially based on amino acid analysis, and subsequently by HPLC. In thelonger term it is likely that only rDNA-derived hGH standards will be widely available, owing to the practicaland health and safety constraints on preparing pituitary-derived material. r-DNA derived standards would alsoassist in meeting new ISO/CEN standards [1] which will require traceability of results in laboratory andpharmaceutical medicine to a defined material.
Ref 1: ISO/CEN Document: in vitro medical devices-measurement of quantities in samples of biological origin -Metrological traceability of values assigned to calibrators and control material ISO/TC 212/ WG2 N65 prEN17511.2000 Other factors affecting assay performance (specificity, hGH binding proteins )
The draft manuscript [Wood, PJ, Personal View: Growth Hormone Measurements-the need for harmonisation”submitted for publication in the Annals of Clinical Biochemistry] was precirculated. hGH in the pituitary existsin a variety of monomeric and oligomeric forms. In serum an even wider variety of molecular heterogeneity isfound owing to protein binding of the monomeric forms and recirculation of tissue derived fragments. An addedcomplexity is that the proportions of the various molecular forms vary in relation to the time after secretoryevents. Different methods of serum hGH measurement, which, in addition to the commonly used immunoassays,include the eluted stain bioassay, 20kDa specific immunoassays, and the Immunofunctional hGH assay, detect adifferent spectrum of hGH molecular forms. hGH binding-proteins in serum interfere in some immunoassays,adding further obstacles to the achievement of between-method agreement.
In discussion, recent data from the UK NEQAS for hGH were presented showing interference from rDNAderived hGH-binding protein 1 in all hGH methods in the scheme, although the extent of interference differedmarkedly between methods Round table discussion: How can we improve the standardisation of hGH assays ?
Choice of calibrant and units
In the medium term the options are to continue to use IS 80/505 (pituitary derived) or to change to IS 98/574 (r-DNA derived). The potential health and safety implications in continuing to use pituitary derived material werenoted. There was a clear consensus that in principle, r-DNA hGH (IS 98/574) should be used as calibrant, andsome manufacturers were already using, or were actively investigating use of its predecessor, IS 88/624.
As the specific activity of IS 98/574 is well defined (3IU/mg ) assay results using this material could be reportedin mass, molar, or bioactivity units. Molar units would be inappropriate in view of the molecular heterogeneity ofhGH in biological fluids. Bioactivity units would maintain continuity with established data, but were lessappropriate where the structure and purity of the primary calibrant could be defined in physico-chemical terms. Itwas agreed that in principle, mass units would be most appropriate. It was noted that the use of mass units wouldalso achieve some uniformity with reporting practice in the US, although the mass content of these earlierstandards had not been accurately defined.
While agreeing in principle with the introduction of 98/574 and use of mass units, manufacturers emphasisedthat implementation of this change through the manufacturing, testing and product regulatory processes, wouldrequire considerable time and resources. Customers would require advance notice of the change, and revision ofreference values and conversion factors would be required. Changes affecting assay results would also need to besubmitted to the FDA for approval in the US market. It was essential therefore that change should beimplemented in a planned and co-ordinated way, with support at an international level. It was proposed that theprofessional bodies should take a leading role in co-ordinating the change.
The timescale agreed for the change was to inform laboratory users and clinicians by Dec 2002, and to introducethe new standards by Dec 2003.
Annual Review for 2000 - Appendix (Growth hormone) Other activities to facilitate the change were discussed.
Dual calibration during the changeover period: Manufacturers could provide both calibrators (80/505 and98/574) during the changeover period to provide customers with some flexibility in managing the transition.
However, this would introduce the potential for error, and might be best reviewed at a later date.
Reference ranges: A systematic review of the literature commissioned by the Royal College of Pathologistsrevealed the paucity of well established hGH reference ranges and responses in stimulation tests, [Butler, J,Biochemical tests of growth hormone status in short children, accepted for publication, Annals of ClinicalBiochemistry, 2000]. In view of the practical and ethical difficulties experienced by manufacturers andlaboratories in establishing such data, it was agreed that a group of clinical and laboratory staff wouldcollaborate in reviewing the data to determine in the first instance, whether improved guidelines for normalresponses could be derived. These could in principle be used to determine guidelines for normal response using98/574 as calibrant. Dr J Seth, Miss J Butler and Prof. M Preece agreed to work on this task.
Communication: Change of calibrant and units would require a major educational effort. The Internet andProfessional Associations would be key routes for communication.
Antibody specificity and other factors affecting inter-method agreement
The literature gave no convincing evidence to support the use of specific 22kDa hGH methods, or methods withbroader specificity in clinical diagnosis, although most of the historical data had been based on broad specificityRIAs. Manufacturers were invited to comment on the rationale for their choice of antibody specificity. Theircomments revealed both theoretical and practical reasons to support the approach they used, whether 22 kDahGH –specific, or broader specificity. They were unanimous in making clear that in the absence of goodevidence favouring one approach to specificity over the other, they would be unable to justify the substantialcosts and labour involved in reformulating their products.
Other factors affecting assay results were discussed. The importance of pre-analytical effects was noted, e.g. theincrease in 20kDa hGH and 20kDa hGH dimers during storage of serum at room temperature.
It was also thought vitally important to understand more clearly the effects of hGH –binding proteins on assayperformance, as these were potentially a major cause of poor inter-method agreement.
Provision of materials to assist harmonisation of hGH assay results.
It was agreed that well characterised preparations of 20kDa hGH and of hGH-binding proteins would berequired by manufacturers and others in developing and characterising methods for hGH measurement. NIBSCwould investigate sources of these with a view to preparing suitable research preparations.
The question was raised as to whether a serum reference panel, containing known concentrations of hGH forms,and analogous to that available for serum steroid assays, would be of value. Although this appeared to be notpractically straightforward, NIBSC would investigate possible approaches.
Other related issues
The meeting had provided an invaluable opportunity for IVD manufacturers, laboratory scientists, clinicians,NIBSC and UK NEQAS to formulate plans for improved assay performance, and the question arose as towhether the format of the meeting could be used for other analytes. It was agreed that such meetings would beuseful, but only where a significant change in practice, (e.g. change in standard) was proposed. Cecelia Brown(BIVDA) commented on the value of direct discussion between those with current practical experience inindustry and in diagnostic laboratories, and stressed the importance of early involvement of such experiencedpractitioners in formulating guidelines.
Annual Review for 2000 - Appendix (Growth hormone) Summary of agreements and actions
1. r-DNA- derived hGH (IS 98/574) should replace the current pituitary- derived standard for hGH (IS 80/505).
2. Immunoassay results using the new standard should be reported in mass 3. The change in calibrant and units must be done in a planned and co- ordinated way, and NIBSC, UK NEQAS and relevant international professional bodies will liaise with manufacturers to manage the change.
4. The change in standardisation should be complete by Dec 2003, with early notice to users being complete by Dec 2002.
5. In order to assist manufacturers in providing guidelines for users, a group of laboratory scientists and clinicians would examine existing datawith a view to providing normal ranges and responses to provocative testsusing the new standard.
6. In the absence of evidence indicating a diagnostic advantage in using22kDa hGH specific or broad specificity assays, no recommendation can bemade at present for standardisation of antibody specificities.
7. The effects of other factors affecting the reliability of hGH results require further investigation. NIBSC would investigate the preparation ofhGH-binding protein 1 and 20kDa hGH as research reagents for use incharacterising assay performance. The preparation of serum referencepreparations would also be investigated.
8. Further experiments will be performed to investigate the effects of hGH-binding proteins on assay performance 9. The views of the meeting would be passed to the WHO Expert Committee on Biological Standardisation, and the ECBS commentsreported back.
John Seth (UK NEQAS, Edinburgh)Adrian Bristow (NIBSC, Potters Bar)Nov 2000 Annual Review for 2000 - Appendix (Growth hormone) Appendix A. Participants
Supplier of International Standards
Laboratory Quality Assurance
Manufacturers
Manufacturers organisations
Laboratory scientists and clinicians
Chairmen
Manufacturers invited but not attending:
Hybritech, IDS, Pharmacia, (withdrawing from hGH manufacturing) Annual Review for 2000 - Appendix (Growth hormone)

Source: http://www.birminghamquality.org.uk/DLopen/nibscgh.pdf