Received 16 September 2009;
accepted 17 September 2009.
Available online 8 October 2009.
Measuring the concentration of growth hormone (GH) in blood samples taken during dynamic tests represents the basis for diagnosis of growth hormone related disorders, namely growth hormone deficiency and growth hormone excess. Today, a wide spectrum of immunoassays are in use, enabling rapid and sensitive determination of growth hormone concentrations in routine diagnostics. From a clinical point of view several difficulties exist with the use and interpretation of GH assay results in the assessment of GH related disorders: Many physiological factors such as fat mass, age and gender influence the outcome of dynamic tests, overall leading to significant inter-individual differences in GH responses. However, in addition to the physiological variability, considerable variability exists in GH assay results obtained by different immunoassays. Unfortunately, all the new technical advances in the field of GH measurement techniques have not reduced this methodological variability. To a large extent, the actual values reported for the GH concentration in a sample depend on the method used by the respective laboratory. Obviously, such discrepancies limit the applicability of consensus guidelines on diagnosis and treatment in clinical practice.
This review summarizes current practices for GH measurement with respect to the methods used, their limitations and the clinical consequences of the existing heterogeneity in GH immunoassay results.
Keywords: Growth hormone; Immunoassays; Standardization
- 1. Methods of GH measurement
- 2. Comparability of GH assay results
- 3. Factors leading to poor comparability of GH assay results
- 3.1. Molecular heterogeneity of the analyte
- 3.2. Reference preparations for GH assays
- 3.3. Impact of GHBP and other matrix components
- 4. Clinical consequences of poor comparability of GH assay results
- 5. Points for discussion
- 5.1. Issues related to the molecular heterogeneity of GH
- 5.1.1. Which GH isoform should be measured?
- 5.1.2. Is it possible to “translate” GH assay results from one assay to another?
- 5.2. Issues related to assay calibration
- 5.3. Issues related to assay validation and performance control
- 5.4. Issues related to the clinical interpretation of GH assay data
1. Methods of GH measurementIn recent decades several methods have been developed for the measurement of growth hormone (GH) activity or concentration in biological fluids. From the very early demonstrations of GH activity in human plasma  and  to the more recent development of bioassays or radioreceptor assays  and , there have been attempts to not only measure the presence (concentration), but also the activity of the GH molecules in a patient’s sample. However, the latter methodologies never reached a distribution among clinical laboratories comparable to that of the immunoassays. The main advantage of immunoassays is that they allow large series of samples to be easily analysed very quickly, an important factor for the clinical laboratory given the increased demand for GH determinations.
In the years following the description of the first radioimmunoassay for insulin , several radioactive GH assays based on the same immunological principle were reported , ,  and . Later, enzyme immunoassays  and other non-radioactive immunoassays  were also developed for the measurement of GH. Today, most of the automated high-throughput analysers in routine use in clinical laboratories utilize chemiluminescence assays to measure GH. In addition to changing the label from radioactivity to other reporting systems, the evolution of GH assays has also been characterized by a switch from competitive immunoassays based on polyclonal antisera to sandwich type immunoassays employing monoclonal antibodies. Competitive assays use labelled GH as a tracer, which competes with the unlabelled GH in the sample for binding to an antibody. The higher the concentration of GH in a sample, the less likely the labelled tracer GH will bind to the antibody. Therefore, there is an inverse correlation between signal intensity and concentration of the analyte. Frequently employing polyclonal antibodies, these assays typically exhibited a lower detection limit in a range between 0.5–1.0 μg/L. In contrast, today’s sandwich type immunoassays for GH employ two antibodies directed against different epitopes on the surface of the GH molecule. One antibody is used to capture the GH molecules in a sample, whereas the second antibody is labelled and used to translate the bound GH molecule into a signal. The intensity of the signal obtained is proportional to the amount of GH present in the samples. Sandwich type assays for GH are available with various labels (radioactivity, enzyme linked and fluorescence), but most automated assay systems currently in use in large laboratories are based on chemiluminescence. An overview on widely used commercially available GH assays is given in Table 1. These immunoassays usually have a sensitivity of around 0.2 μg/L or better, and the most sensitive ones have been reported to reach a sensitivity below 0.002 μg/L .
Characteristics of commonly used commercial assays for GH (according to manufacturers instructions/kit inserts available to the authors or according to published data). Calibration has changed for several assays in recent years, and the process is ongoing. Thus, the information provided might not be up to date for all assay lots on the market, and also not for all countries. Furthermore, the list of assays is not complete. Additional hGH assays exist, including an unknown number of in-house assays.
Recommended sample material
Immulite 2000/2500 hGH
Automated immunometric assay (mAb + pAb), chemiluminescence
Calibrator changed to 98/574 since lot 206 in most European countries. Calibrator 80/505 in the US
Automated immunometric assay (mAb + pAb), chemiluminescence
Access ultrasensitive hGH
Automated immunometric assay (pAb + mAb), chemiluminescence
20 kD < 4%
Calibrator 80/505 according to 
Automated immunometric assay (mAb + mAb), chemiluminescence
20 kD < 2%
Serum, plasma (heparin, citrate)
Information from manufacturers flyer, assay available 9/2009
Immunometric assay (mAb + mAb), time-resolved fluorescence
22 kD specific
Serum, plasma (heparin)
Immunometric assay (mAb + mAb), radioactive
Serum, plasma (heparin, EDTA)
Immunometric assay, radioactive
Provides standard concentrations for both calibrators
Immunometric assay (mAb + mAb), radioactive
20 kD < 5%
DSL-10-1900 ACTIVE hGH ELISA
Immunometric assay, enzyme linked
Provides standard concentrations for both calibrators
Abbreviations: pAb, polyclonal antiserum; mAb, monoclonal antibody.
View Within Article
It is important to keep in mind that many of the important studies on diagnosis and treatment of GH related diseases in the past have reported GH levels measured by competitive assays based on polyclonal antisera. The outcome of these studies of course still influences the current recommendations and guidelines for diagnosis and treatment of GH related diseases , ,  and . In addition, the actual values of GH concentrations mentioned in the older publications sometimes still influence clinical practice, although the assay methods underlying these numbers (or cut off values) are not used anymore . It will be described below how changes in assay methods over time have altered the actual values of GH concentrations reported by the various assays, and why these differences create a continued problem for the clinical and scientific community.
2. Comparability of GH assay results
For many years external quality assessment schemes (EQAS) have documented large method- or laboratory dependent discrepancies between immunoassay results for several steroid- and peptide hormones. From this perspective, GH assays are only one unpleasant example of the difficulties in standardizing peptide hormone measurements  and . The problem has been reported for GH assays from studies in several countries in the past , ,  and . The variability in concentrations reported for a single sample by different methods exceeds 100% in many cases. Obviously, this leads to problems when results from one study are compared to those from another study which used a different assay to measure GH concentrations. More relevant in clinical practice, this makes it impossible to compare GH assay results from one hospital to another when different assays and laboratories are used. Some studies directly compared results from different assays in the same clinical samples in one laboratory. In addition to confirming disagreement between results from different GH assays these studies also illustrated that the assay discrepancies affect the clinical interpretation of the results ,  and .
3. Factors leading to poor comparability of GH assay results
Before discussing causes of method dependent discrepancies it is important to mention that in the case of GH the impact of problems in the preanalytical period on the results of laboratory measurements can be considered negligible. Although concentrations fluctuate considerably in the human body, GH exhibits a remarkable stability in vitro. It has been demonstrated that GH concentrations in serum samples remain stable for at least 24 h at room temperature and for even longer periods at 4 °C  and .
3.1. Molecular heterogeneity of the analyte
A major contributor to the heterogeneity of GH assay results is the fact that GH as secreted by the pituitary gland and circulating in the human body is a mixture of several molecular isoforms rather than a homogenous substance  and . The most abundant isoform is a 22 kD GH molecule, followed by a smaller 20 kD molecule derived from alternative splicing. Hetero- and homodimers of the GH isoforms exist, and also multimers are present in human serum . GH shares this feature of “molecular heterogeneity” with several other proteohormones such as hCG – and GH assays unfortunately share the poor comparability between methods with assays for other proteohormones  and . Given the molecular heterogeneity of the analyte, measurement of such analytes by immunoassays inherently will lead to discrepant results depending on the antibodies used: Each immunoassay will pick up a selected spectrum of isoforms only – the epitopes of the respective antibodies used in the assays define which fraction of isoforms are translated into a signal in the respective assay. Because polyclonal antisera represent a mixture of different antibodies (each of which can potentially recognize different epitopes), the spectrum of isoforms measured by assays involving polyclonal antisera is most likely quite broad. In contrast, assays involving only monoclonal antibodies (which target only a very specific epitope) tend to recognize only a one or a few of the isoforms present in a sample. As a consequence, assays involving polyclonal antisera usually give higher results than those based on monoclonal assays. In addition, the heterogeneity of the analyte also serves as an explanation for why the agreement between GH assays was better when the less specific, polyclonal antisera were used  and , compared to recent years when monoclonal antibody based assays became predominant . The higher specificity of the monoclonal antibodies led to more pronounced differences in the measurement results of different GH assays, and the overall between-method variability in the UK national EQAS worsened during the 90’s from about 17% to about 30%.
For most of the widely used GH assays information on the specific GH isoform measured by the respective assay is not provided by the manufacturer. One can assume that most of the assays recognize a certain spectrum of isoforms, and almost all will bind 22 kD GH because it is part of the antigen used by the manufacturers to generate the antibodies for the assays. Not relevant to clinical routine diagnostics, but helpful for research applications, are various GH assays designed to exclusively or preferentially measure individual GH isoforms. Such assays exist for the measurement of the 20 kD isoform  and , for placental GH or GH-V , for the whole spectrum of “non-22 kD” isoforms  and for the determination of the relative abundance of the 22 kD isoform .
A very recent difficulty in using GH measurements occurred when the GH receptor antagonist Pegvisomant was introduced in clinical practice for the treatment of acromegaly . This drug is a modified GH molecule, leading to blockade of the GH receptor. The drug circulates at concentrations 100–1000 times higher than endogenous GH. Because of the molecular similarity to wild type GH and because of the high concentration, most commercially available GH assays cross-react with Pegvisomant . Interestingly, the bias in GH assay results can be both, positive and negative, depending on the assay used: If both antibodies in a sandwich type GH assay cross-react with Pegvisomant, the results will be extremely high, whereas when only one of the antibodies cross-reacts, GH assay results can be falsely low. Only certain assays are available to specifically measure only endogenous GH in the presence of Pegvisomant , ,  and , and the clinical relevance of GH determinations during Pegvisomant treatment is under investigation.
3.2. Reference preparations for GH assays
Another factor obviously influencing the comparability of assay results is the reference preparation or “standard” used to calibrate the assay. The concentration of an analyte in an immunoassay generally is determined by comparing the signal generated in the sample (with unknown concentration) to the signals generated in standard samples with known amounts of the analyte (standard curve). Therefore, immunoassay measurements are relative measurements in nature, and the composition and quality of the respective reference preparation has tremendous impact on the concentration reported by the assay. Historically, the first reference preparations for GH assays were pituitary extracts. Accordingly, the first international reference preparation (IRP) 66/217, introduced in 1969, and the following preparation 80/505, introduced in 1982, both contained a variety of GH isoforms, although some kind of purification was applied to increase the relative percentage of the 22 kD isoform. The “true” GH content in these two preparations was unknown and they were arbitrarily assigned concentrations of 2.0 and 2.6 U/mg, respectively. Subsequently, reference preparations of recombinant origin and consisting of the 22 kD isoform exclusively (IRP 88/624 and 98/574) became available. Although the recombinant origin and the high purity would allow expression of the GH content in mass units, the preparation IRP 88/624 was assigned to a biopotency of 3.0 U/mg to allow some kind of comparison to the previous pituitary extracts. The concomitant use of reference preparations of pituitary and recombinant origin in different GH assays clearly adds to the discrepancies between the assay results. In addition, the use of two units (mU/L and μg/L) and the adoption of a variety of conversion factors between the units further impaired the comparability of GH assay results  and . Obviously, the change in the standard preparation has major influence on the absolute concentrations reported by different assays .
3.3. Impact of GHBP and other matrix components
In circulation, up to 50% of GH is complexed with a high affinity GH binding protein (GHBP) . GHBP corresponds to the extracellular domain of the GH receptor, and its concentration varies with nutritional and metabolic conditions . The presence of GHBP in serum samples becomes important for GH assays when epitopes of the respective antibodies used in a GH assay are no longer accessible due to steric hindrance by GHBP. Such a situation might lead to the underestimation of the GH concentration by the respective assay in samples containing GHBP at high concentrations. This problem was less pronounced in the past when assays with polyclonal antisera and long incubation times were used , but seems to be very relevant for newer assays using monoclonal antibodies with a defined epitope in conjunction with a very short incubation time. These assays have been shown to be susceptible to interference from GHBP, with a negative bias approaching 50% in some assays for GHBP concentrations still in the physiological range ,  and .
Little is known about the potential impact of other matrix components on the results of GH assays. However, as some manufacturers, for example, do not recommend the use of EDTA plasma for use in GH assays because results tend to be higher than results from serum samples, one might speculate about factors other than GHBP influencing the assay results.
4. Clinical consequences of poor comparability of GH assay results
Poor comparability of GH assays has had significant clinical consequences, particularly for the diagnosis and monitoring of patients with GH deficiency and GH excess, acromegaly. The principal consequence is that no uniform GH criteria for these disorders can be developed so long as the assays remain heterogeneous.
In the case of acromegaly, determination of generalizable criteria for glucose-suppressed GH levels, obtained during an OGTT, to diagnose or determine cure of the disease has not been possible. With current sensitive and specific GH assays it is clear that the criteria in use in the past with polyclonal radio-immunoassays (RIA) of 2.0 μg/L for OGTT nadir or 2.5 μg/L for mean GH levels are no longer valid. With assays utilizing monoclonal antibodies it became evident that normal GH suppression after glucose is actually much <1 μg/L  and  and that levels should be <1 μg/L to exclude acromegaly or establish its remission  and . However, exactly where below 1 μg/L this cut off should be cannot be uniformly agreed upon because the data vary depending upon the GH assay used ,  and . A number of studies support a cut off of 1 μg/L ,  A.M. Arafat, M. Mohlig, M.O. Weickert, F.H. Perschel, J. Purschwitz, J. Spranger, C.J. Strasburger, C. Schofl and A.F. Pfeiffer, Growth hormone response during oral glucose tolerance test: the impact of assay method on the estimation of reference values in patients with acromegaly and in healthy controls, and the role of gender, age, and body mass index, J. Clin. Endocrinol. Metab. 93 (2008), pp. 1254–1262. View Record in Scopus | Cited By in Scopus (13),  and  while others utilizing different assays suggest cut offs of 0.5 μg/L  and  or 0.25 μg/L . An OGTT cut off of 0.3 μg/L has also been suggested with an assay that utilizes 2 monoclonal antibodies that are 22 K GH specific and standards that are calibrated to the recombinant human GH 22 K specific reference preparation 88/624 . One source of variability in these cut offs is the GH standard utilized. The GH standards used in recent studies assessing normal and acromegaly OGTT criteria have varied from the WHO standard 80/505 which contains both 20 K and 22 K GH ,  and  to polyclonal GH standards obtained from other sources such as the NIH and National Hormone and Pituitary Program  and  and to rhGH standards 88/624  or 98/574 . However, even when the same group of acromegaly samples were run in two different assays calibrated to the rhGH standard (98/574) significant differences were demonstrated . In this study, the mean GH nadir in controlled acromegaly patients was 0.98 ± 0.26 μg/L with the Immulite 2000 assay vs. 0.5 ± 0.15 μg/L with the Nichols Advantage assay and in patients with active disease, these were 7.98 ± 1.7 μg/L and 4.5 ± 1.2 μg/L with these two assays, respectively. Nadir GH levels in healthy subjects are similarly variable making it difficult to establish normative data. When OGTT samples in healthy subjects were measured with three different assays calibrated to the rhGH standard (98/574) mean nadir GH concentrations varied from 0.13 ± 0.01 μg/L (range 0.05–0.99) with the Immulite 2000 assay to 0.06 ± 0.005 μg/L (range 0.02–0.5 μg/L) with the Nichols assay and to 0.015 ± 0.002 μg/L(range 0.00066–0.25 μg/L) with the Diagnostic Systems Laboratories Elisa assay. Differences in antibodies and other factors as described in this review likely play roles in determining the observed differences in GH data. Thus, it is clear that an optimal cut off applicable to all laboratories cannot be provided because of assay differences.
Variability between GH assays has led similarly to difficulty in standardizing those criteria for provocative testing of GH that can differentiate GH sufficiency from insufficiency. One study compared 699 peak GH levels obtained during stimulation testing of children undergoing evaluations for GH deficiency that were assayed in three different reference assays and local laboratories. The study found large differences between assays in mean GH levels, which ranged from 5.4 mU/l to 10.3 mU/l. Furthermore, diagnostic assignment varied in up to 27% of the children depending on the assay used . Another study also found significant assay-dependent differences in peak stimulated GH in children with GH deficiency . In this study and another, categorization of children into GH sufficient vs. deficient groups also varied markedly between assays  and . Some authors have advocated for centralized reassessment of GH peak sera in reference centers in order to reduce variability in GH testing and utilization of these centralized results for treatment decisions . Similar differences are likely to exist when measuring GH levels obtained during stimulation testing for the determination of GH deficiency in adults.
Therefore, it is clear that determination of optimal GH criteria for both acromegaly and GH deficiency has been hindered by the current heterogeneity of GH assays. A further consequence of the inability to derive uniform criteria is that published criteria developed with one assay are often adopted to interpret GH levels measured with another in the research setting, in guidelines publications and in clinical practice. In the latter setting, the use of inaccurate criteria to guide clinical decision-making could, in some cases, lead to errors in clinical care. The adoption of criteria developed with one assay to use with another may also have actually hindered the development of more uniformity because of a false sense of security that these criteria could be utilized more generally. Until such a time when more uniformity of GH assay methodology is adopted and/or appropriate normative data for each assay are developed, published diagnostic GH criteria for acromegaly or GH deficiency can only serve as a general guideline. Some steps could improve more widespread use of published criteria in the meantime as outlined below.
5. Points for discussion
There is increasing awareness of the problem of GH assay variability  and . Recently an international collaborative started to discuss options to improve the situation  and . and the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) has put the issue of GH assays on the agenda by implementing a Working Group on GH assay standardization  and . Experience with attempts to standardize other hormone assays has demonstrated that standardization of immunoassays requires compromises – no one approach is likely to adequately address all the issues related to the problem. The quality required for a particular assay has to consider clinical needs and theoretical and practical limitations . Future attempts to standardize or harmonize GH assays should consider the following issues (key points summarized in Table 2):
Suggested points for the discussion of future attempts to standardize GH assays.
Points to discuss
Molecular heterogeneity of GH
Which isoforms are important to be measured?
Is a mathematical conversion of concentration values obtained by different assays possible and reliable?
How can the use of a single calibrant (recombinant hGH 98/574) be achieved worldwide?
Which factors have to be taken into account when preparing standards for assays?
What is a “commutable” standard preparation?
Assay validation and performance control
What recommendations can be made for procedures to investigate isoform crossreactivity and potential impact of GHBP?
Is there a need to eliminate GHBP interference?
Is it required to determine the respective epitopes of all antibodies used in GH assays?
Clinical interpretation and cut off values
Are “assay specific cut off values” an attainable goal?
What could be done to establish cut off values applicable for a broad spectrum of assays?
Which physiological factors have to be taken into account for the establishment of cut off values for dynamic tests?
View Within Article
5.1. Issues related to the molecular heterogeneity of GH
5.1.1. Which GH isoform should be measured?
It is a matter of debate whether a more specific or a more permissive recognition of GH isoforms is preferable from a clinical point of view. It has been shown convincingly that different isoforms are biologically active , and therefore one is potentially missing information with assays that measure only one isoform . However, one might also argue that even up to today no specific clinical condition has been found where the response to dynamic tests (stimulation or suppression) differed with respect to the different molecular isoforms. As 22 kD GH is by far the most abundant isoform, it might be sufficient to have a specific measurement of this isoform only. Whatever the final recommendation with respect to the preferred analyte for GH assays – it is very important for each GH assay’s proper interpretation that each assay manufacturer provide specific information on the isoform measured by their respective GH assay.
5.1.2. Is it possible to “translate” GH assay results from one assay to another?
The discrepancy between the concentrations obtained from measurement of GH by two different immunoassays depends on factors related to the respective assays (antibody specificity, assay design, etc.), but also on factors related to the individual sample analysed, mainly the relative abundance of the GH isoforms present in the respective sample. Therefore, it is impossible to make results from different immunoassays comparable by simply applying a linear “conversion factor”. This is especially important to keep in mind when considering assays employing monoclonal antibodies – they are particularly sensitive to the isoform composition of each sample. Furthermore, the respective impact of GHBPs present in the individual samples will affect the “commutability” of assay results. From this point of view it is unlikely that published recommendations for cut off values based on older, polyclonal GH assays can be “translated” into cut offs valid for the newer, automated immunoassay systems. An experimental approach to reduce discrepancies between GH assay results has been reported recently : So-called “harmonization samples” (representing “average patient samples”) are distributed among laboratories, measured with the local assays and used to correct the respective GH assay results for all other samples measured in the lab. The authors of that study report a reduction in between-laboratory variability for GH assay results from 15% to <7%. The concept is interesting, but also has its problems: First, it is unlikely that an “average patient sample” is able to adequately represent the potential interference of isoforms, GHBP and other factors. Furthermore, the study had been conducted with a rather small number of laboratories, and it might be difficult to apply this concept to a larger international setting because the availability of sufficient amounts of adequate harmonization samples is a prerequisite.
5.2. Issues related to assay calibration
5.2.1. Which calibrator should be used for GH assays?
The international collaborative on GH assay standardization has recommended the sole use of the recombinant IRP of 22 kD GH 98/574, and that GH assay results be expressed in mass units (μg/L) of this preparation. Although there are still some assays on the market that are calibrated against the older pituitary extract preparations, there is a tendency towards standardization on this point. The study by Tanaka et al.  has clearly shown the potential benefit of such harmonization of the assay calibrant. Furthermore, only reference materials of recombinant origin and high purity could allow traceability of GH assay results . Meanwhile, some important journals in the field have decided to only publish papers that include GH data that are expressed in mass units of the most recent International Standard 98/574  and . However, this approach is problematic as several manufacturers – including the market leader – have yet to implement the change in the calibrator in all countries. Thus, GH assays from the same manufacturer have been – and still are – available with different calibrators in different European countries. If and when the corresponding change will be implemented for assays available in the US is still unclear. As a result, the journal’s requirement at this time that GH values be expressed only in mass units of the IRP 98/574 is not tenable. However, it is clear that researchers must be extremely careful in reporting the correct calibration for the GH assays they used to measure their samples.
Clearly, several problems with GH assays cannot be solved solely by using a single calibrator for all assays. Furthermore, it is important to keep in mind that the potency of standard preparations used for GH assays also can be affected by the way the reference preparations (e.g. 98/574) are treated to obtain the calibration solution used for the respective assay. E.g. the matrix of the assay specific calibrators is different between assays from different manufacturers. Commutability of immunoassay results is a complex task . It remains to be investigated to what degree the “standardization of the standard” alone will lead to better comparability of GH assay results .
5.3. Issues related to assay validation and performance control
5.3.1. What information must be given for GH assays with respect to specificity and interference?
Precise characterization of each GH assay with respect to both the GH isoform spectrum recognized, and to potential interference from GHBP is desirable. However, conducting the necessary experiments is challenging because of the limited availability of the substances required. Although recombinant 22 kD GH is easily available, access to the 20 kD isoform is more difficult, and general availability of homo- and heterodimers of isoforms might remain impossible. For GHBP, some recombinant preparations are commercially available, but these preparations are produced in non-mammalian cell culture systems. As a consequence, the recombinant GHBP preparations contain non-glycosylated GHBP only, whereas endogenous GHBP is glycosylated. Whether or not glycosylation status affects the degree of potential interference of GHBP in immunoassays is unknown. The scientific community should provide clear guidelines on reagents and procedures to be used in GH assay validation.
5.4. Issues related to the clinical interpretation of GH assay data
5.4.1. What are quality criteria for establishing cut off values?
Given the heterogeneity of GH assay results, most guidelines emphasize the need for “method specific cut off values” and point out the problem of assay variability , ,  and . Such an approach may be ideal as some aspects of each GH assay will never be completely universally comparable. However, for most of the assays currently in use in large clinical laboratories world wide there are no published studies or reference data available. Establishing such data requires resources and sometimes is very difficult or even impossible for ethical reasons – e.g. in children. In addition, normative data for GH will need to individualized for age, gender and BMI as these factors are important for determining GH levels in OGTT testing as well as stimulation testing for GH. Laboratories working in a highly competitive economic environment very likely are unable to produce such reference values for any new GH assay introduced. One strategy would be to develop reference populations from a multi-center effort that could then be shared across laboratories and assayed with multiple assays, but this too would require considerable resources.
5.4.2. Given the inherent differences between concentrations determined by different immunoassays – what can be done to improve reliability of applying guidelines and recommendations?
If assay specific normative data cannot be established then certain minimal steps may help establish some comparability and allow for applicability of cut offs with one assay to the other. The first and most straightforward step is universal adoption of the rhGH standard as described above. A second step could be to advocate for transparency on the part of assay manufacturers about the characteristics on their assay, in particular the specificity of the antibodies and the potential interferences as outlined throughout this manuscript. Collaborative efforts between researchers and industry to establish reference criteria for different categories of GH assays along with transparency of the factors described above could help create some published criteria that are more generalizable.
 C.A. Gemzell, F. Heijkenskjold and L. Strom, A method for demonstrating growth hormone activity in human plasma, J. Clin. Endocrinol. Metab. 15 (1955), pp. 537–546. View Record in Scopus | Cited By in Scopus (2)
 M.T. Dattani, P.C. Hindmarsh, C.G. Brook, I.C. Robinson, T. Weir and N.J. Marshall, Enhancement of growth hormone bioactivity by zinc in the eluted stain assay system, Endocrinology 133 (1993), pp. 2803–2808. View Record in Scopus | Cited By in Scopus (22)
 M.A. Lesniak, P. Gorden, J. Roth and J.R. Gavin 3rd, Binding of 125I-human growth hormone to specific receptors in human cultured lymphocytes. Characterization of the interaction and a sensitive radioreceptor assay, J. Biol. Chem. 249 (1974), pp. 1661–1667. View Record in Scopus | Cited By in Scopus (43)
 R.S. Yalow and S.A. Berson, Assay of plasma insulin in human subjects by immunological methods, Nature 184 (Suppl. 21) (1959), pp. 1648–1649.
 D.S. Schalch and M.L. Parker, A Sensitive double antibody immunoassay for human growth hormone in plasma, Nature 203 (1964), pp. 1141–1142.
 H.G. Morris, Y. Arai, C.J. Hlad Jr., R. Tompkins and H. Elrick, Rapid quantitative immunologic assay for human growth hormone, J. Clin. Endocrinol. Metab. 24 (1964), pp. 417–424.
 H. Orskov, H.G. Thomsen and H. Yde, Wick chromatography for rapid and reliable immunoassay of insulin, glucagon and growth hormone, Nature 219 (1968), pp. 193–195. View Record in Scopus | Cited By in Scopus (114)
 A.H. Wu, A selected history and future of immunoassay development and applications in clinical chemistry, Clin. Chim. Acta 369 (2006), pp. 119–124. Abstract | Article | PDF (519 K) | View Record in Scopus | Cited By in Scopus (14)
 I.M. Chapman, M.L. Hartman, M. Straume, M.L. Johnson, J.D. Veldhuis and M.O. Thorner, Enhanced sensitivity growth hormone (GH) chemiluminescence assay reveals lower postglucose nadir GH concentrations in men than women, J. Clin. Endocrinol. Metab. 78 (1994), pp. 1312–1319. View Record in Scopus | Cited By in Scopus (135)
 Consensus guidelines for the diagnosis and treatment of adults with growth hormone deficiency: summary statement of the growth hormone research society workshop on adult growth hormone deficiency, J. Clin. Endocrinol. Metab. 83 (1998) 379–381.
 Consensus guidelines for the diagnosis and treatment of growth hormone (GH) deficiency in childhood and adolescence: summary statement of the GH research society, J. Clin. Endocrinol. Metab. 85 (2000) 3990–3993.
 A. Giustina, A. Barkan, F.F. Casanueva, F. Cavagnini, L. Frohman, K. Ho, J. Veldhuis, J. Wass, K. Von Werder and S. Melmed, Criteria for cure of acromegaly: a consensus statement, J. Clin. Endocrinol. Metab. 85 (2000), pp. 526–529. View Record in Scopus | Cited By in Scopus (514)
 S. Melmed, F. Casanueva, F. Cavagnini, P. Chanson, L.A. Frohman, R. Gaillard, E. Ghigo, K. Ho, P. Jaquet, D. Kleinberg, S. Lamberts, E. Laws, G. Lombardi, M.C. Sheppard, M. Thorner, M.L. Vance, J.A. Wass and A. Giustina, Consensus statement: medical management of acromegaly, Eur. J. Endocrinol. 153 (2005), pp. 737–740. View Record in Scopus | Cited By in Scopus (87)
 B.P. Hauffa, N. Lehmann, M. Bettendorf, O. Mehls, H.G. Dorr, C.J. Partsch, H.P. Schwarz, N. Stahnke, H. Steinkamp, E. Said, S. Sander and M.B. Ranke, Central reassessment of GH concentrations measured at local treatment centers in children with impaired growth: consequences for patient management, Eur. J. Endocrinol. 150 (2004), pp. 291–297. View Record in Scopus | Cited By in Scopus (26)
 F. Bidlingmaier, W.J. Geilenkeuser, R. Kruse and G. Rohle, Our experience with quality control in current growth hormone assays, Horm. Res. 36 (Suppl. 1) (1991), pp. 1–4. View Record in Scopus | Cited By in Scopus (8)
 J. Seth, A. Ellis and R. Al-Sadie, Serum growth hormone measurements in clinical practice: an audit of performance from the UK National External Quality Assessment scheme, Horm. Res. 51 (1999), pp. 13–19. View Record in Scopus | Cited By in Scopus (29)
 P. Morsky, U. Tiikkainen, A. Ruokonen and H. Markkanen, Problematic determination of serum growth hormone: experience from external quality assurance surveys 1998–2003, Scand. J. Clin. Lab. Invest. 65 (2005), pp. 377–386. View Record in Scopus | Cited By in Scopus (7)
 T. Tanaka, K. Tachibana, A. Shimatsu, N. Katsumata, T. Tsushima, N. Hizuka, K. Fujieda, S. Yokoya and M. Irie, A nationwide attempt to standardize growth hormone assays, Horm. Res. 64 (Suppl. 2) (2005).
 A.C. Celniker, A.B. Chen, R.M. Wert Jr. and B.M. Sherman, Variability in the quantitation of circulating growth hormone using commercial immunoassays, J. Clin. Endocrinol. Metab. 68 (1989), pp. 469–476. View Record in Scopus | Cited By in Scopus (93)
 M.L. Granada, A. Sanmarti, A. Lucas, I. Salinas, A. Carrascosa and M. Foz, Assay-dependent results of immunoassayable spontaneous 24-hour growth hormone secretion in short children, Acta Paediatr. Scand. (Suppl. 370) (1990), pp. 63–70 (discussion 71). View Record in Scopus | Cited By in Scopus (40)
 H. Markkanen, T. Pekkarinen, M.J. Valimaki, H. Alfthan, R. Kauppinen-Makelin, T. Sane and U.H. Stenman, Effect of sex and assay method on serum concentrations of growth hormone in patients with acromegaly and in healthy controls, Clin. Chem. 52 (2006), pp. 468–473. View Record in Scopus | Cited By in Scopus (15)
 R.L. Derr, S.J. Cameron and S.H. Golden, Pre-analytic considerations for the proper assessment of hormones of the hypothalamic–pituitary axis in epidemiological research, Eur. J. Epidemiol. 21 (2006), pp. 217–226. View Record in Scopus | Cited By in Scopus (4)
 M.J. Evans, J.H. Livesey, M.J. Ellis and T.G. Yandle, Effect of anticoagulants and storage temperatures on stability of plasma and serum hormones, Clin. Biochem. 34 (2001), pp. 107–112. Abstract | Article | PDF (70 K) | View Record in Scopus | Cited By in Scopus (70)
 G. Baumann, M.W. Stolar and T.A. Buchanan, The metabolic clearance, distribution, and degradation of dimeric and monomeric growth hormone (GH): implications for the pattern of circulating GH forms, Endocrinology 119 (1986), pp. 1497–1501. View Record in Scopus | Cited By in Scopus (23)
 C.M. Sturgeon, P. Berger, J.M. Bidart, S. Birken, C. Burns, R.J. Norman and U.H. Stenman, Differences in recognition of the 1st WHO international reference reagents for hCG-related isoforms by diagnostic immunoassays for human chorionic gonadotropin, Clin. Chem. 55 (2009), pp. 1484–1491. View Record in Scopus | Cited By in Scopus (5)
 C.M. Sturgeon and A.R. Ellis, Standardization of FSH, LH and hCG – current position and future prospects, Mol. Cell. Endocrinol. 260–262 (2007), pp. 301–309. Abstract | Article | PDF (276 K) | View Record in Scopus | Cited By in Scopus (11)
 P. Chatelain, B. Bouillat, R. Cohen, G. Sassolas, J.C. Souberbielle, A. Ruitton, M.O. Joly and J.C. Job, Assay of growth hormone levels in human plasma using commercial kits: analysis of some factors influencing the results, Acta Paediatr. Scand. (Suppl. 370) (1990), pp. 56–61 (discussion 62). View Record in Scopus | Cited By in Scopus (30)
 E.O. Reiter, A.H. Morris, M.H. MacGillivray and D. Weber, Variable estimates of serum growth hormone concentrations by different radioassay systems, J. Clin. Endocrinol. Metab. 66 (1988), pp. 68–71. View Record in Scopus | Cited By in Scopus (44)
 T. Tsushima, Y. Katoh, Y. Miyachi, K. Chihara, A. Teramoto, M. Irie and Y. Hashimoto, Serum concentration of 20K human growth hormone (20K hGH) measured by a specific enzyme-linked immunosorbent assay. Study group of 20K hGH, J. Clin. Endocrinol. Metab. 84 (1999), pp. 317–322 ISSN: 0021-972x. View Record in Scopus | Cited By in Scopus (46)
 K.C. Leung, C. Howe, L.Y. Gui, G. Trout, J.D. Veldhuis and K.K. Ho, Physiological and pharmacological regulation of 20-kDa growth hormone, Am. J. Physiol. Endocrinol. Metab. 283 (2002), pp. E836–E843. View Record in Scopus | Cited By in Scopus (25)
 Z. Wu, M. Bidlingmaier, R. Dall and C.J. Strasburger, Detection of doping with human growth hormone, Lancet 353 (1999), p. 895. Abstract | Article | PDF (43 K) | View Record in Scopus | Cited By in Scopus (93)
 C.L. Boguszewski, M.C. Boguszewski, F. de Zegher, B. Carlsson and L.M. Carlsson, Growth hormone isoforms in newborns and postpartum women, Eur. J. Endocrinol. 142 (2000), pp. 353–358. View Record in Scopus | Cited By in Scopus (9)
 M. Bidlingmaier, J. Suhr, A. Ernst, Z. Wu, A. Keller, C.J. Strasburger and A. Bergmann, High-sensitivity chemiluminescence immunoassays for detection of growth hormone doping in sports, Clin. Chem. 55 (2009), pp. 445–453. View Record in Scopus | Cited By in Scopus (9)
 P.J. Trainer, W.M. Drake, L. Katznelson, P.U. Freda, V. Herman-Bonert, A.J. van der Lely, E.V. Dimaraki, P.M. Stewart, K.E. Friend, M.L. Vance, G.M. Besser, J.A. Scarlett, M.O. Thorner, C. Parkinson, A. Klibanski, J.S. Powell, A.L. Barkan, M.C. Sheppard, M. Malsonado, D.R. Rose, D.R. Clemmons, G. Johannsson, B.A. Bengtsson, S. Stavrou, D.L. Kleinberg, D.M. Cook, L.S. Phillips, M. Bidlingmaier, C.J. Strasburger, S. Hackett, K. Zib, W.F. Bennett and R.J. Davis, Treatment of acromegaly with the growth hormone-receptor antagonist pegvisomant, N. Engl. J. Med. 342 (2000), pp. 1171–1177. View Record in Scopus | Cited By in Scopus (317)
 A.N. Paisley, K. Hayden, A. Ellis, J. Anderson, G. Wieringa and P.J. Trainer, Pegvisomant interference in GH assays results in underestimation of GH levels, Eur. J. Endocrinol. 156 (2007), pp. 315–319. View Record in Scopus | Cited By in Scopus (11)
 M.O. Thorner, C.J. Strasburger, Z. Wu, M. Straume, M. Bidlingmaier, S.S. Pezzoli, K. Zib, J.C. Scarlett and W.F. Bennett, Growth hormone (GH) receptor blockade with a PEG-modified GH (B2036-PEG) lowers serum insulin-like growth factor-I but does not acutely stimulate serum GH, J. Clin. Endocrinol. Metab. 84 (1999), pp. 2098–2103. View Record in Scopus | Cited By in Scopus (75)
 J.D. Veldhuis, M. Bidlingmaier, S.M. Anderson, Z. Wu and C.J. Strasburger, Lowering total plasma insulin-like growth factor I concentrations by way of a novel, potent, and selective growth hormone (GH) receptor antagonist, pegvisomant (B2036-peg), augments the amplitude of GH secretory bursts and elevates basal/nonpulsatile GH release in healthy women and men, J. Clin. Endocrinol. Metab. 86 (2001), pp. 3304–3310. View Record in Scopus | Cited By in Scopus (52)
 A. Colao, R. Pivonello, R.S. Auriemma, M.C. De Martino, M. Bidlingmaier, F. Briganti, F. Tortora, P. Burman, I.A. Kourides, C.J. Strasburger and G. Lombardi, Efficacy of 12-month treatment with the GH receptor antagonist pegvisomant in patients with acromegaly resistant to long-term, high-dose somatostatin analog treatment: effect on IGF-I levels, tumor mass, hypertension and glucose tolerance, Eur. J. Endocrinol. 154 (2006), pp. 467–477. View Record in Scopus | Cited By in Scopus (48)
 S.R. Brian, M. Bidlingmaier, M.P. Wajnrajch, S.A. Weinzimer and S.E. Inzucchi, Treatment of acromegaly with pegvisomant during pregnancy: maternal and fetal effects, J. Clin. Endocrinol. Metab. 92 (2007), pp. 3374–3377. View Record in Scopus | Cited By in Scopus (9)
 A. Pokrajac, G. Wark, A.R. Ellis, J. Wear, G.E. Wieringa and P.J. Trainer, Variation in GH and IGF-I assays limits the applicability of international consensus criteria to local practice, Clin. Endocrinol. 67 (2007), pp. 65–70. View Record in Scopus | Cited By in Scopus (26)
 P.J. Trainer, J. Barth, C. Sturgeon and G. Wieringaon, Consensus statement on the standardisation of GH assays, Eur. J. Endocrinol. 155 (2006), pp. 1–2. View Record in Scopus | Cited By in Scopus (0)
 T. Amit, M.B. Youdim and Z. Hochberg, Clinical review 112: does serum growth hormone (GH) binding protein reflect human GH receptor function?, J. Clin. Endocrinol. Metab. 85 (2000), pp. 927–932. View Record in Scopus | Cited By in Scopus (49)
 T. Jan, M.A. Shaw and G. Baumann, Effects of growth hormone-binding proteins on serum growth hormone measurements, J. Clin. Endocrinol. Metab. 72 (1991), pp. 387–391. View Record in Scopus | Cited By in Scopus (26)
 L. Ebdrup, S. Fisker, H.H. Sorensen, M.B. Ranke and H. Orskov, Variety in growth hormone determinations due to use of different immunoassays and to the interference of growth hormone-binding protein, Horm. Res. 51 (1999), pp. 20–26. View Record in Scopus | Cited By in Scopus (31)
 C. Jansson, C. Boguszewski, S. Rosberg, L. Carlsson and K. Albertsson-Wikland, Growth hormone (GH) assays: influence of standard preparations, GH isoforms, assay characteristics, and GH-binding protein, Clin. Chem. 43 (1997), pp. 950–956. View Record in Scopus | Cited By in Scopus (55)
 T.K. Hansen, S. Fisker, B. Hansen, H.H. Sorensen, J.S. Christiansen, J.O. Jorgensen and H. Orskov, Impact of GHBP interference on estimates of GH and GH pharmacokinetics, Clin. Endocrinol. 57 (2002), pp. 779–786. View Record in Scopus | Cited By in Scopus (18)
 P.U. Freda, K.D. Post, J.S. Powell and S.L. Wardlaw, Evaluation of disease status with sensitive measures of growth hormone secretion in 60 postoperative patients with acromegaly, J. Clin. Endocrinol. Metab. 83 (1998), pp. 3808–3816. View Record in Scopus | Cited By in Scopus (127)
 A.C. Costa, A. Rossi, C.E. Martinelli Jr., H.R. Machado and A.C. Moreira, Assessment of disease activity in treated acromegalic patients using a sensitive GH Assay: should we achieve strict normal GH levels for a biochemical cure?, J. Clin. Endocrinol. Metab. 87 (2002), pp. 3142–3147. View Record in Scopus | Cited By in Scopus (41)
 P.U. Freda, Current concepts in the biochemical assessment of the patient with acromegaly, Growth Horm. IGF Res. 13 (2003), pp. 171–184. Abstract | Article | PDF (176 K) | View Record in Scopus | Cited By in Scopus (36)
 N. Hattori, A. Shimatsu, Y. Kato, H. Koshiyama, Y. Ishikawa, H. Assadian, T. Tanoh, M. Nagao and H. Imura, Growth hormone responses to oral glucose loading measured by highly sensitive enzyme immunoassay in normal subjects and patients with glucose intolerance and acromegaly, J. Clin. Endocrinol. Metab. 70 (1990), pp. 771–776. View Record in Scopus | Cited By in Scopus (30)
 A.M. Arafat, M. Mohlig, M.O. Weickert, F.H. Perschel, J. Purschwitz, J. Spranger, C.J. Strasburger, C. Schofl and A.F. Pfeiffer, Growth hormone response during oral glucose tolerance test: the impact of assay method on the estimation of reference values in patients with acromegaly and in healthy controls, and the role of gender, age, and body mass index, J. Clin. Endocrinol. Metab. 93 (2008), pp. 1254–1262. View Record in Scopus | Cited By in Scopus (13)
 L. De Marinis, A. Mancini, A. Bianchi, R. Gentilella, D. Valle, A. Giampietro, P. Zuppi, C. Anile, G. Maira and A. Giustina, Preoperative growth hormone response to thyrotropin-releasing hormone and oral glucose tolerance test in acromegaly: a retrospective evaluation of 50 patients, Metabolism 51 (2002), pp. 616–621. Abstract | PDF (90 K) | View Record in Scopus | Cited By in Scopus (12)
 M. Mercado, A.L. Espinosa de los Monteros, E. Sosa, S. Cheng, V. Mendoza, I. Hernandez, C. Sandoval, G. Guinto and M. Molina, Clinical-biochemical correlations in acromegaly at diagnosis and the real prevalence of biochemically discordant disease, Horm. Res. 62 (2004), pp. 293–299. View Record in Scopus | Cited By in Scopus (6)
 R.A. Feelders, M. Bidlingmaier, C.J. Strasburger, J.A. Janssen, P. Uitterlinden, L.J. Hofland, S.W. Lamberts, A.J. van der Lely and W.W. de Herder, Postoperative evaluation of patients with acromegaly: clinical significance and timing of oral glucose tolerance testing and measurement of (free) insulin-like growth factor I, acid-labile subunit, and growth hormone-binding protein levels, J. Clin. Endocrinol. Metab. 90 (2005), pp. 6480–6489. View Record in Scopus | Cited By in Scopus (18)
 O. Serri, C. Beauregard and J. Hardy, Long-term biochemical status and disease-related morbidity in 53 postoperative patients with acromegaly, J. Clin. Endocrinol. Metab. 89 (2004), pp. 658–661. View Record in Scopus | Cited By in Scopus (38)
 E.V. Dimaraki, C.A. Jaffe, R. DeMott-Friberg, W.F. Chandler and A.L. Barkan, Acromegaly with apparently normal GH secretion: implications for diagnosis and follow-up, J. Clin. Endocrinol. Metab. 87 (2002), pp. 3537–3542. View Record in Scopus | Cited By in Scopus (98)
 Y. Rakover, I. Lavi, R. Masalah, T. Issam, E. Weiner and I. Ben-Shlomo, Comparison between four immunoassays for growth hormone (GH) measurement as guides to clinical decisions following GH provocative tests, J. Pediatr. Endocrinol. Metab. 13 (2000), pp. 637–643 (In Process Citation). View Record in Scopus | Cited By in Scopus (10)
 A. Juul, S. Bernasconi, P.E. Clayton, W. Kiess and S. DeMuinck-Keizer Schrama, European audit of current practice in diagnosis and treatment of childhood growth hormone deficiency, Horm. Res. 58 (2002), pp. 233–241. View Record in Scopus | Cited By in Scopus (23)
 P.J. Trainer, J. Barth, C. Sturgeon and G. Wieringa, Consensus statement on the standardisation of GH assays, Eur. J. Endocrinol. 155 (2006), pp. 1–2. View Record in Scopus | Cited By in Scopus (30)
 M. Bidlingmaier and C.J. Strasburger, What endocrinologists should know about growth hormone measurements, Endocrinol. Metab. Clin. North Am. 36 (2007), pp. 101–108. Abstract | Article | PDF (118 K) | View Record in Scopus | Cited By in Scopus (4)
 U.H. Stenman, Immunoassay standardization: is it possible, who is responsible, who is capable?, Clin. Chem. 47 (2001), pp. 815–820.
 M. Hayakawa, Y. Shimazaki, T. Tsushima, Y. Kato, K. Takano, K. Chihara, A. Shimatsu and M. Irie, Metabolic effects of 20-kilodalton human growth hormone (20K-hGH) for adults with growth hormone deficiency: results of an exploratory uncontrolled multicenter clinical trial of 20K-hGH, J. Clin. Endocrinol. Metab. 89 (2004), pp. 1562–1571. View Record in Scopus | Cited By in Scopus (10)
 H.A. Ross, Reporting growth hormone assay results in terms of one consensus recombinant standard preparation offers less than optimal reduction of between-method variation, Clin. Chem. Lab. Med. 46 (2008), pp. 1334–1335. View Record in Scopus | Cited By in Scopus (3)
 M.E. Molitch, D.R. Clemmons, S. Malozowski, G.R. Merriam, S.M. Shalet, M.L. Vance and P.A. Stephens, Evaluation and treatment of adult growth hormone deficiency: an Endocrine Society Clinical Practice Guideline, J. Clin. Endocrinol. Metab. 91 (2006), pp. 1621–1634. View Record in Scopus | Cited By in Scopus (110)
 K. Iwahara, C. Tanabe and M. Maekawa, Comparison of access ultrasensitive human growth hormone assay to monoclonal antibody-based immunoradiometric assay, Clin. Chim. Acta 376 (2007), pp. 248–249. Abstract | Article | PDF (145 K) | View Record in Scopus | Cited By in Scopus (1)
Corresponding author. Address: Endocrine Research Laboratories, Medizinische Klinik – Innenstadt, Ludwig-Maximilians University, Ziemssenstr. 1, Munich, Germany. Tel.: +49 8951602277.
1 MB has received lecture fees from DiaSorin, consultancy fees from IDS and research support from IDS, Diasorin, Siemens and DSL.