Christie Hospital National Health Service Trust, Withington, Manchester M20 4BX, United Kingdom
Address all correspondence and requests for reprints to: S. M. Shalet, M.D., Christie Hospital National Health Service Trust, Wilmslow Road, Withington, Manchester M20 4BX, United Kingdom. E-mail: stephen.m.shalet{at}man.ac.uk.
Untreated acromegaly is associated with increased morbidity and mortality; however, when disease activity is controlled, the relative mortality risk is reduced toward normal (1). Thus, treatment of the disease is indicated in almost all cases once the diagnosis is established. Optimization of treatment strategies is still evolving with the introduction of new medical approaches, such as long-acting somatostatin analogs and a GH receptor antagonist, and the less frequent use of conventional pituitary radiotherapy (2); nonetheless, for the majority of endocrinologists, pituitary surgery performed by a specialist pituitary surgeon remains the initial treatment of choice for most, but not all, newly diagnosed patients with acromegaly.
A number of controversies in the field of acromegaly remain, one of which is how to define biochemically that disease activity is controlled. Problems of interpretation have arisen because of assay changes and a relative lack of age-matched normative data using modern assays. Thus, GH can now be measured by two-site monoclonal antibodies and nonisotopic assays with enhanced sensitivity rather than polyclonal RIA of limited sensitivity; it is not possible, however, to apply a simple conversion factor to compare GH levels measured by current assays with those obtained in the past by RIA.
In parallel with our greater ability to measure lower circulating levels of GH, IGF-I measurement has been used increasingly in the diagnosis and management of acromegaly. This is despite the fact that there are still problems with current IGF-I assays, including a relative lack of age-matched normative data, lack of standardization, susceptibility to interference from binding proteins, and the lack of a pure international reference preparation.
The serum IGF-I level, however, reflects GH status and correlates better with clinical disease activity in acromegaly than GH measurements. In addition, the observation by Swearingen et al. (3) that normalization of the serum IGF-I level in patients operated on for acromegaly was associated with normalization of the standardized mortality ratio further supported the greater weight placed on IGF-I measurement.
Therefore, how exactly should we monitor disease activity biochemically after pituitary surgery for acromegaly? IGF-I measurements alone? IGF-I and GH measurements? The proponents of only measuring IGF-I might argue that: 1) the IGF-I level correlates better than GH levels with clinical disease activity (4); 2) achieving a normal IGF-I level predicts normalization of life expectancy; and 3) a single IGF-I estimation is easier to perform than GH levels measured throughout an oral glucose tolerance test (OGTT) or a GH day curve.
The counter argument for measuring both IGF-I and GH levels is based on the observation that: 1) discordant values for GH and IGF-I estimations may be seen in up to 30% of patients (5); 2) in the presence of discordant GH and IGF-I values, significant morbidity may be present; 3) the prognostic value in terms of mortality risk of a normal postoperative IGF-I level is based on very few studies; and 4) there is a nonlinear relationship between GH and IGF-I levels (6) such that a marked reduction, but clear failure, to normalize previously very high GH levels in a surgically treated acromegalic may be associated with little change in the IGF-I level; thus, biochemical audit of what actually took place at surgery requires knowledge of GH levels as well as IGF-I status. Furthermore, there is the belief, albeit unproven thus far, that significant debulking of a GH-secreting pituitary adenoma with suboptimal reduction of GH levels increases the subsequent capacity of certain (i.e. somatostatin analogs), but not all other, modalities of treatment to normalize GHIGF-I status.
In this issue of JCEM, Freda et al. (7) provide an additional argument (i.e. the prediction of recurrence) for measuring both GH and IGF-I levels in postoperative patients with acromegaly; the authors had already established previously that nadir GH levels (measured in an immunoradiometric assay with a sensitivity of 0.05 µg/liter) after a 100-g oral glucose load were all less than 0.14 µg/liter in healthy adults (8). In the current study of 110 postoperative patients with acromegaly evaluated by OGTT, 76 were in remission, defined by a normal IGF-I; of these patients, 50 had normal nadir GH (<0.14 µg/liter; group 1) and 26 had abnormal nadir GH (>0.14 µg/liter; group 2) levels. Of the 76 patients in remission, 49 (30 subjects from group 1 and 19 from group 2) underwent serial longitudinal oral glucose tolerance testing every 12 yr over a mean follow-up time of 3.2 yr (range, 16.5); the initial pattern of GH suppression persisted in most patients. IGF-I levels remained normal in all patients in group 1, but five of the 19 patients in group 2 developed an elevated IGF-I level and, thus, a biochemical recurrence (7).
The subtle impairment of GH suppression observed in remission group 2 was reproducible over time, suggesting that this represents persistent GH dysregulation in these patients. In addition, the abnormal nadir GH levels in group 2 patients were accompanied by other evidence of greater GH secretion than that seen in postoperative patients with normal GH suppression (group 1). Preceding the biochemical recurrence, IGF-I levels were in the upper half of the normal range in all five patients and in the upper quartile in three of the five patients (7).
In the light of these findings, we need to consider several questions: What is the explanation for the coexistence of GH dysregulation and a normal IGF-I level? What are the practical implications of these findings in terms of clinical management?
Gender, age, and prior radiotherapy are all known to alter the relationship between GH and IGF-I in acromegaly (9, 10), with women and the elderly having lower IGF-I values for a given GH level; none of the five patients with a biochemical recurrence in this study had received radiotherapy. The explanation for the discordancy between GH and IGF-I status may lie in the definition of a normal IGF-I level; even with a large normative database, the range of decade-based normal values for IGF-I is wide. The existence of a much narrower individual normal range for a hormone measurement, compared with the group reference range used to define normality, is established for T3 and T4 levels (11). Therefore, it is possible that the five patients with a biochemical recurrence had a normal IGF-I level defined by the cross-sectional normative data but a pathologically elevated IGF-I level defined by their own personal normal range. There are several strands of information that are not inconsistent with this speculation. First, all five patients with acromegaly had an IGF-I level in the upper half of the normal range in the period preceding biochemical recurrence (7). Second, more than 50% of adult patients with severe GH deficiency have a normal IGF-I level but nearly always a level that is in the lower half of the normal age-adjusted range (12). It is tempting to speculate that in the majority of the adult GH-deficient patients with the lower half of the normal range IGF-I levels, IGF-I levels were in the upper half of the normal range before the onset of their pituitary disease, and that a trend in the opposite direction existed for the five postoperative patients with acromegaly who went on to develop a biochemical recurrence (7).
The practical issue arising from these observations of Freda et al. (7) pertains to the schedule and nature of the patient follow-up postoperatively. Is it now possible to predict more accurately in which patient a recurrence is more likely to occur? There have been many studies of recurrence rates after transsphenoidal surgery for acromegaly, but even recent figures reported have varied considerably (019%), possibly due to a lack of consistency in the criteria used to define remission (3, 13, 14, 15). Previous studies almost certainly overestimated the recurrence rate by including patients who actually had persistent postoperative disease not detected by less sensitive GH assays. Distinguishing recurrence from persistent postoperative disease is not resolved by the current study (7), in which the mean time from surgery to the initial OGTT evaluation in group 2 patients was 4.8 yr, because normal OGTT nadir GH values were never established at any time during the postoperative period in the five patients in whom a biochemical recurrence (pathologically elevated IGF-I level) occurred (7).
The authors, quite reasonably, also caution against over-interpreting their findings, because the follow-up period for some of their patients is still relatively short; thus, information derived from long-term follow-up of their remission cohorts is required before drawing firm conclusions about the true accuracy of the OGTT nadir GH level to predict either long-term remission or increased risk of recurrence after surgery for acromegaly.
If, however, long-term follow-up data support their preliminary conclusions (7), then we shall need to consider whether patients with evidence of GH dysregulation and a normal IGF-I level are to be followed up more intensively. What exactly will be the frequency of this increased monitoring? Twice a year instead of once a year? Once a year instead of every two years? Furthermore, how much anxiety will be induced in the patient by knowledge that they are at an increased risk of disease recurrence?
We also need to know more about any morbidity, disease related, experienced postoperatively by patients treated for acromegaly, who exhibit GH dysregulation and a normal IGF-I level. If present, such patients may benefit from medical therapy, but at this subtle end of the range of disease activity, careful monitoring of changes in the IGF-I level will be needed to avoid inducing the biological sequelae of GH deficiency.
The authors have, however, succeeded in tilting the debate about the nature of biochemical monitoring of patients operated on for acromegaly a little further toward GH and IGF-I measurements rather than IGF-I estimation alone.
Footnotes
Abbreviation: OGTT, Oral glucose tolerance test.
Received November 14, 2003.
Accepted November 14, 2003.
References
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |