What is Active Acromegaly and Which Parameters do we Have?

S. J. C. M. M. Neggers; N. R. Biermasz; A. J. van der Lely


Clin Endocrinol. 2012;76(5):609-614. 

In This Article

Biochemical Markers, GH and IGF1

Assessment of GH, and sex and age-matched IGF1 concentrations are important biochemical parameters for the diagnosis of acromegaly, and at present are the most accepted measures to monitor treatment response. Suppression of both GH[2–4] and IGF1 concentrations[5–7,11] to current valid criteria for cure, that is, normal IGF1 and GH < 2·5 μg/l have been associated with improved "normalised" mortality.

As mortality is a hard end-point, the consensus statements seem to focus mainly on normalization of GH and IGF1.[8–10] In these statements, the biochemical target ranges for GH concentration have changed over time, but this is not the case for IGF1. Controlled IGF1, in these statements, means an IGF1 within the sex- and age-adjusted normal limits. For GH, it is different. According to the earlier criteria, the mean of a 24-h GH profile should be to <2·5 μg/l or a GH nadir after glucose of <1·0 μg/l.[8] Using these limits, Dekkers et al.[11] demonstrated that the standardized mortality rate (SMR) is still slightly increased at 1·09. The SMR from this meta-analysis is based mainly upon studies conducted in patients with acromegaly treated with transsphenoidal surgery. With current effective medical treatments, the SMR outcome could be different.

An important issue concerns the assays used to assess GH and IGF1, as these have changed over the years. Currently, ultra-sensitive assays are used to assess GH. Holdaway et al.[12] have demonstrated in a New Zealand cohort of patients with acromegaly that a single GH of <1·0 μg/l was associated with normalization of mortality. More recent consensus statements introduced new limits to reflect these and other findings (single GH 1·0 μg/l or a GH nadir of <0·4 μg/l).[13] So, when the original Cortina criteria[8] are applied, some patients still have active disease as the measured GH level is perceived to have decreased because of the current use of more sensitive immunoradiometricassays rather than the previous, less sensitive, polyclonal radioimmunoassays. Consequently, the puzzle is not solved by just lowering GH cut-offs. With the more sensitive GH assays, there is a lack of adequate gender-specific normative data, standardization and assay validation.[14,15] Depending on which commercial assay is used, the means of the GH nadirs of healthy volunteers differ from 0·13 to 0·015 μg/l. The explanation for the heterogeneity between results of the sensitive assays are variable calibration, epitope specificity of the antibody and variable specificity of antibody recognition of different circulating GH isoforms in the serum.[15] Finally, the lack of adequate standardization in the use of mass units and international units and the variety of conversion factors lead to marked problems of interpretation.[15]

The situation is improving, however. A recent report provides consensus statements on the standardization and evaluation of GH and IGF1 assays. This concludes that major improvements are necessary in the areas of assay performance and comparability. The group of involved experts recommended that a commutable standard for each assay be implemented for worldwide use and that its recommendations be applied to accomplish the task of providing reliable and clinically useful results.[16]

The problems for IGF1 are not that different. Circulating serum levels of IGF1 are influenced by age, gender, season, nutritional status and concomitant disease such as diabetes mellitus. Serum IGF1 is mainly bound to IGF1 binding proteins and acid-labile subunit: only approximately 1% is unbound.[17] To assess IGF1 serum levels, these binding proteins need to be displaced. One of the most common methods is functional displacement by IGF2.[18]

In diabetes, IGF1 can be glycosylated at the recognition site of the antibody of the IGF1 assay.[19] Also, increased proteolytic activity of IGF1 binding protein 3[20] can influence the assay's results.[20] Acromegaly and diabetes mellitus are associated; therefore, this could influence the assessed IGF1 results in our patients.

The assay calibration by the International Reference Reagent (IRR 87/518) is biased because of uncertain purity compared with the available recombinant IGF1 preparation.[21] Additionally, there is a lack of adequate normative data for the commercially available IGF1 assays.

The variability in assay performance for GH and IGF1, coupled with the use of inappropriate conversion factors and normative data, undermines the applicability of the consensus criteria to local practice.[22]

Practical Aspects

After surgery, GH and IGF1 assessment are time-dependent. A long half-life and other factors regulating IGF1 may result in its elevation several months after surgery,[13,23] whereas GH during oGTT is an early predictor of remission or failure.[13,23]

During long-acting somatostatin analogue (LA-SRIF) treatment, GH nadir assessment cannot be performed after glucose loading.[14,24] Therefore, IGF1 is considered to be the most feasible assessment. However, if patients still have symptoms of active acromegaly, mean or single GH measurements can still be helpful.[13] During pegvisomant (PEG-V) treatment, IGF1 is the only reliable assessment. PEG-V mimics GH and therefore interferes with most commercially available GH assays making results invalid.[25]

So, in everyday practice at the outpatient clinic, IGF1 is the most feasible parameter to assess. Therefore, the aim is to normalize serum IGF1 levels via a reduction of GH action or GH overproduction.[1,8,9,13,26] A tangible advantage of this approach is that the efficacy of different treatments can easily be compared by means of serum IGF1 measurements. This is more practicable than either time-consuming, and costly, 24 h GH measurements[26] or assessment of GH nadir after glucose loading, which is only valid after surgery and not during treatment with LA-SRIF analogues.[24] This also applies to comparisons between the effects of long-acting LA-SRIF analogue therapies, PEG-V and surgery.

By this, we assume that serum IGF1 levels adequately and uniformly reflect disease activity. However, it can be argued that this assumption is not valid. Acromegaly patients with normal IGF1 during LA-SRIF treatment still had elevated nadir GH levels as compared with patients with a normal IGF1 after surgery.[27] The LA-SRIF treated patients also had a reduced disease-specific health status compared with the surgically cured patients.[27] The combination of an elevated IGF1 level and a normal GH level is sometimes seen after radiation therapy because radiotherapy causes a flat GH secretory pattern.[28] On the other hand, several factors have been identified that can result in lower IGF1 relative to GH levels, such as chronic inflammatory disorders and anorexia nervosa (which can impair IGF1 production by the liver), nutritional or gastrointestinal disorders such as hepatic or renal failure, poorly controlled type 1 diabetes, oral oestrogens and hypothyroidism.[14,26] Acromegaly patients with long-term "remission" may have some signs of GH excess. This can be observed as arthralgia, mild hypertension, relative glucose intolerance and numbness of the hands.[29–31] By antagonizing this excess of GH, QoL improves.[30]

There is also evidence from animal models for this hypothesis of mild GH excess. In mice, GH can have temporal and tissue-specific effects that are independent of IGF1 serum levels. In adipose tissue, kidney and skeletal muscle, local activity of GH can be blocked with relatively low doses of PEG-V that do not change serum IGF1.[32] For example, glomerulosclerosis is caused by high GH levels not by raised IGF1. This was confirmed in mice 20 years ago.[33,34]

Finally, in a prospective, double-blind, placebo-controlled, crossover trial of PEG-V treatment in acromegaly patients with "normal" IGF1, QoL was assessed by a disease-specific QoL questionnaire, Acromegaly Quality of Life Questionnaire (AcroQoL)[35,36] and a disease-specific symptom questionnaire (PASQ).[25,37] After 16 weeks of treatment with 40 mg PEG-V weekly, QoL improved, as indicated by an increased AcroQoL score total and AcroQoL score physical dimension. The magnitude of the improvement in AcroQoL score of 6·4% was equal to the observations of Paisley et al.[38] when elevated IGF-I was reduced to the age-adjusted normal range. This was accompanied by a reduction in symptoms assessed by PASQ, (the total PASQ score and the single PASQ questions, perspiration, soft tissue swelling and overall health status). The same symptoms in the PASQ scores, perspiration and soft tissue swelling, also decrease after PEG-V treatment in previously uncontrolled patients. This seems to occur without a significant correlation with change in IGF1 levels.[39] This clinically relevant improvement also failed to be detected by other currently available biochemical markers.[30] Although auto or paracrine GH or IGF1 effects may explain some of these findings, there is evidence for a new concept called "extra hepatic acromegaly".[26]

Normalization of levels of total serum IGF1 and GH do not necessarily reflect optimal QoL, nor relief of symptoms in patients with acromegaly.[5,26,30,40–43] From the patient's perspective, QoL is one of the most important parameters of disease control. So the goals of treatment are to normalize IGF1 and GH levels and activity, and improve QoL.

There is a group of patients with active disease who do not have impaired QoL. It is not logical that in these patients, QoL is a good bioassay or could improve more. However, there are a limited number of patients without any symptoms or decreased QoL otherwise they would not have sought medical care.

As briefly discussed previously, QoL can be assessed by different questionnaires. There are more general questionnaires, which can be used to study general health concepts and functional status in different situations and there are disease-specific questionnaires. To quantify perceived health, QoL and symptoms in patients with acromegaly, PASQ[25,37] and the AcroQoL are available and are described here in more detail:


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: