Quality assurance in the analysis of growth hormone and insulin-like growth factor I in disorders of the somatotropic axis

Introduction

The growth hormone (GH) is an anabolic proteohormone. It is produced in the anterior pituitary and secreted under the control of the hypothalamic hormones growth-hormone-releasing hormone (GHRH) and somatostatin [1]. Insulin-like growth factor I (IGF-I) is mainly formed under the control of GH in the liver and secreted. In plasma, IGF-I is bound to specific carrier proteins (IGF-binding protein IGFBP). GH and IGF-I primarily affect the growth and differentiation of cells. In childhood and adolescence, they contribute significantly to longitudinal growth; in adulthood, they regulate, inter alia, the glucose and lipid metabolism [1, 2]. A reliable laboratory analysis and informed interpretation of measurement results are the cornerstones of an accurate diagnosis and the treatment to be derived from this for patients with disorders of growth and GH secretion.

However, assay-specific problems and insufficiently characterized reference ranges create doubts about the possibility of the clinical implementation of current consensus recommendations on the diagnosis and treatment of GH deficiency and acromegaly [3–5]. This leads to the need to achieve harmonization of GH and IGF-I measurements as well as consistent quality assurance. Such efforts were already made in the past. Among them, one should point out the results of the interdisciplinary consensus conference in Keswick (Virginia, USA) in 2009, which was attended by representatives of international professional associations and representatives of regulatory agencies, manufacturers of reagents and the pharmaceutical industry [6].

The following provides a summary of the results of a workshop attended by clinical endocrinologists and laboratory physicians in collaboration with the Academy for Education and Training of the German Society of Endocrinology and the Endocrinological Laboratory Diagnostics section of the German Society for Clinical Chemistry and Laboratory Medicine. The objective of the workshop was to determine the level of quality in the diagnosis of GH and IGF-I and to develop ways for improvement.

Recommendations for the standardization of assays

The central problem of current GH and IGF-I assays is the poor comparability of measurement results obtained by different analytical methods. Currently, the use of different reference preparations to calibrate assays contributes significantly to the high variability of measurement results. The exclusive use of the recombinant international reference preparation 98/574 as a single calibrator has been demanded for GH for some time [6]. As for IGF-I assays, the Keswick consensus speaks clearly in favor of the recombinant WHO standard IS 02/254 [6]. In fact, it has been shown that the uniform use of a recombinant calibrator renders measurement results obtained by different methods more comparable [7].

In serum, GH occurs in different isoforms, which are detected variously by the current antibodies in commercial assays. A standardized GH assay should be specific to the 22-kDa form and provide accurate measurement results, especially in the lower concentration range [6]. The dissociation constants and binding epitopes of the antibodies used should also be specified, like the cross-reactivity of the antibodies with other GH isoforms. As a standard preparation and antibodies have the highest impact on the variability, it seems desirable over the long term to use only assays for routine tests that are identical in terms of the standard and antibodies.

The clinical relevance of analytical interference with GH binding proteins (GHBP) [8] is questionable. However, manufacturers should check the interference of their GH assays with GHBP and provide a physiological account. In regard to the interference of IGF binding proteins with IGF-I assays [9], the respective method used to separate IGF-I from its high-affinity binding proteins should be described and validated. The impact on the analytical method stemming from altered binding protein patterns in connection with comorbidities such as renal failure, liver dysfunction or diabetes mellitus should be known as well [6]. When studying the effect of the GH receptor antagonist pegvisomant, a structural analog of the human GH used for the treatment of acromegaly, GH assays should be used that do not interfere with pegvisomant [10].

Depending on the assay, measurement results of GH and IGF-I assays have so far been stated not only as mass concentrations (μg/L), but also in international units (U/L). But the units were defined arbitrarily, and there is no clear link to the mass of hormones. Concentrations should, therefore, be stated uniformly as mass concentrations [6].

The validation of immunoassays should be done using sera from healthy volunteers and patients with disorders of GH secretion. As part of the validation of IGF-I assays, particularly sera from patients with diabetes mellitus, chronic renal and hepatic impairment should be examined. It may help to establish a serum database that manufacturers can use to run longitudinal tests on, and adjust, the performance of their assays. The assay quality (accuracy and precision) should be described in the lower and upper limits of the reference ranges (IGF-I) or at clinically important decision limits (GH stimulation and suppression tests).

Manufacturers should regularly publish data on batch consistency and inter-assay variability, as well as reports on the quality control of modifications to essential components of assays.

Quality criteria for reference ranges and limits

Normative data must be based on large, clinically relevant reference populations and – depending on the parameters – stratified by age group, gender, or other important factors. Manufacturers should provide assay-specific normative data for IGF-I and for dynamic GH tests, while scientific societies and associations should act as technical advisors to relevant studies.

Given the significant method-based differences in assay results as well as in terms of the expenditure, which good reference range studies entail, a possible harmonization of assay results through mathematical-statistical methods of conversion has been discussed over and over again. The consensus conference [6] and the workshop participants agreed that this did not represent a satisfactory option for IGF-I measurements and the corresponding reference ranges. The participants considered the conversion of limits of dynamic tests justified only when ethical reasons made it impossible to conduct studies to determine method-specific decision limits.

Also needed is a better description and improved comparability of reference cohorts with respect to the subsequent patient group, e.g., in connection with phenotypes, drugs and stages of puberty. When using data from epidemiological approaches, the effect of possible influencing factors must be clarified. Gender-specific reference ranges for IGF-I are especially necessary in childhood and puberty after the age of 10 [6]. For the age range between 0 and 20 years, IGF-I reference ranges should be specified for 1-year age intervals. The number of samples should be sufficiently high (n>120 for individual age strata). Longitudinal cross-sectional data are necessary when it comes to puberty. Especially in the case of pubescent adolescents, it is crucial to carry out a differentiated evaluation of measured IGF-I levels. As age and pubertal development correlate only to a very limited degree, the IGF-I reference ranges during this period, if possible, should be adjusted according to the stages of puberty. For adults, increments of five, perhaps even 10, years are sufficient. Ethnicity and body mass index (BMI) do not affect IGF-I measurements significantly, which means that it does not appear to be necessary to adjust reference ranges for these factors [6].

Diagnosis of GH deficiency: decision limits and interpretation of results

The clinical context must be taken into account in connection with the diagnosis and progress-monitoring of treatment of growth and GH-secretion disorders. Diagnostic accuracy correlates with the clinical probability of the presence of the disease to be analyzed. Determination of individual GH levels is not sufficient to confirm GH deficiency, because the hormone is released from the pituitary gland in a pulsatile manner – generally, it cannot be detected between pulses physiologically speaking. In order to assess the functionality of the GH-IGF-I axis reliably, stimulation tests are required. The insulin tolerance test (ITT) is used to check the integrity of the entire hypothalamic-pituitary-adrenal axis. In healthy individuals, insulin-induced hypoglycemia produces a maximum stress response with secretion of GH. The secretion response does not occur in patients with GH deficiency. In Germany, the GHRH-arginine test is frequently used, in addition to the ITT, as a test for the GH stimulation at the central level. It checks the maximum secretory capacity of the hypothalamus and the pituitary gland [3].

Conclusions

The comparability of currently available commercial assays for the determination of GH and IGF-I is still insufficient today. Therefore, it is necessary to raise awareness about the quality of immunoassays, make rational use of diagnostic tools, and counteract an interpretation of measurement results that is often too uncritical. The primary objective should be to make the quality criteria for assays known and thus motivate the manufacturers of diagnostic tools to do more for their implementation. Quality criteria and guidelines should be evaluated regularly to ensure a sustainable quality assurance process. Manufacturers, as well as experts from relevant professional associations, should also plan and conduct interdisciplinary studies to ensure an independent validation of assays and the establishment of assay-specific decision limits and/or reference intervals for appropriate reference populations. Finally, physicians should know about the quality and assays of laboratories and take into account the clinical context of the individual patient when it comes to diagnostic strategy, treatment decisions and follow-ups.