| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Departments of Endocrinology (G.A.K., A.M.I., D.F., P.J.T., C.C.-H., J.P.J., S.L.C., J.P.M., G.M.B., A.B.G.) and Neurosurgery (F.A., I.S.), St. Bartholomew’s and the Royal London Hospitals, London ECIA 7BE, United Kingdom
Address correspondence and requests for reprints to: Prof. A. B. Grossman, Department of Endocrinology, St. Bartholomew’s Hospital, London ECIA 7BE, United Kingdom. E-mail: A.B.Grossman{at}mds.qmw.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
2 ng/mL) after
treatment. Transsphenoidal surgical resection is the best available
treatment for restoring GH to such "safe" levels; however, criteria
for the assessment of the response to treatment are not uniform. To
determine the clinically most useful method of assessing disease
activity postoperatively and identify predictors of a favorable
response to surgical treatment, we have analyzed 67 patients with
acromegaly who underwent transsphenoidal surgery between 1993 and 1998.
We used three different definitions of a satisfactory or safe response:
1) a postoperative mean GH less than 5 mU/L obtained from averaging
five serum GH values obtained throughout one day; 2) a random single GH
less than 5 mU/L; or 3) a serum insulin-like growth factor I (IGF-I)
level within the normal range. Relying on a single GH measurement
alone, 9 of the 23 patients with a single postoperative mean GH level
less than 5 mU/L obtained at least one GH value of more than 5 mU/L
(false positive rate, 28%) and 8 of the patients with a postoperative
mean GH value of more than 5 mU/L obtained a single GH value of less
than 5 mU/L (false negative rate, 15%). Postoperatively, a significant
increase in the fluctuation of random GH values around the mean was
observed in patients who were rendered safe (coefficient of variation,
from 26 ± 2% to 53 ± 6%; P < 0.001)
compared with patients with persistence of inadequately controlled
disease. However, 13% (3 of 23) of patients with mean postoperative GH
levels of less than 5 mU/L had elevated serum IGF-I levels
postoperatively, and 17% (8 of 44) of patients with mean serum GH
levels more than 5 mU/L had postoperative IGF-I levels within the
normal range. There was no difference in the rate of agreement between
mean GH less than 5 mU/L and normalization of IGF-I in relation to the
interval since operation when IGF-I levels were measured. Preoperative tumor size and pretreatment mean GH levels were the major determinants of the outcome of surgery, as patients who were rendered safe had significantly lower preoperative mean GH levels than patients who were not cured (median, 31 mU/L vs. 78.5 mU/L, P < 0.01). IGF-I levels were weakly correlated with tumor size and could not be used to predict the patients who would be rendered safe. Preoperative PRL levels were higher in patients who failed to achieve a surgical satisfactory outcome [498 mU/L (187–857) vs. 196 mU/L (136–315), P < 0.01].
In summary, although single random GH values and IGF-I values are both significantly correlated with mean GH levels, they should not be used as an alternative to averaging several GH values to assess disease activity, because of the pulsatile nature of GH secretion and the multiple factors that may influence serum IGF-I. Because significant discrepancies occur, particularly postoperatively, mean GH levels remain the more reliable indicator of surgical outcome and disease activity. As there is considerably more evidence relating long-term prognosis to serum GH levels than to IGF-I and discrepancies occur between GH levels and IGF-I, we suggest that mean serum GH levels and single IGF-I levels, measured early in the postoperative period, are currently the best biochemical guide to the adequacy of surgery and, hence, the need for further treatment.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Whereas transsphenoidal surgery (TSS) is currently the optimal therapy in terms of the achievement of immediate remission and mean GH levels less than 5 mU/L in a significant number of patients, the majority will require further treatment to maintain GH levels in or near the desirable range (10, 11). Although radiotherapy has been used as an adjunct to surgical treatment, it results in a relatively slow fall of GH levels over the subsequent years, and interim medical treatment with somatostatin analogs, dopamine agonists, or a GH receptor antagonist is usually needed (12). It is, therefore, important to establish reliable predictors of the outcome of TSS and identify pre- and postoperatively who will most probably require further intensive treatment. A number of factors have been suggested as being useful predictors of the success of surgery: these include tumor size (13), preoperative basal GH levels (14), and impaired pituitary reserve, all of which are associated with poor outcome and the need for further treatment after surgery. However, in some of these studies, single random or basal GH values were used as an indicator of disease activity and the response to surgery (13, 14, 15, 16, 17, 18). Although a single random serum GH sample is easier to obtain in practice, normalized mortality has been described only with mean GH levels less than 5 mU/L from multiple sampling throughout the day (5, 6). Over the last decade, estimation of IGF-I as a marker of integrated GH secretion has been suggested by some to be of greater validity than immunoreactive GH levels in the assessment of outcome, in that it should reflect overall GH bioactivity (13, 19, 20, 21, 22, 23, 24). However, we have little data on mortality related to IGF-I (7). Although a number of studies evaluated serum IGF levels in the postoperative follow-up in acromegaly, most of these studies used different criteria to assess residual disease, mainly random GH values or the nadir during glucose suppression, as well as a wide range of different cut-off values (1–5 µg/L). Furthermore, in most of these studies the normal reference range for IGF-I values was not age adjusted. Lack of homogeneity in these studies led to contradictory results in the rate of concordance between postoperative IGF-I and GH criteria for remission (25). In general, the most common discrepancy is the finding of ‘normal’ GH with elevated IGF-I levels; less frequent is the opposite finding of elevated GH with normal IGF levels (14, 15, 18, 19, 26, 27, 28, 29, 30). Recent data provided evidence that the extent of GH suppression with oral glucose should be reconsidered, because more than 50% of patients with active disease (elevated IGF-I) have a nadir GH value less than 1 µg/L when measured with enhanced sensitivity assays (27). On the contrary, very few data have compared pre- and postoperative changes in mean GH levels with IGF-I using as cut-off a mean GH less than 5 mU/L, which is the only value significantly associated with normalized mortality (7, 19, 24, 26). Finally, many factors other than GH—including nutritional state, liver function, serum protease activity, IGF-I-binding proteins, and sex hormones—contribute to the determination of serum IGF-I, and particularly in the adult IGF-I levels may not be directly related to GH (26, 31, 32). Thus, for example, in the adult GH deficiency syndrome at least 30% of patients have normal serum IGF-I levels (33).
We have now evaluated 67 patients with newly diagnosed acromegaly who underwent TSS between 1993 and 1998 to determine: 1) preoperative factors, including IGF-I measurements, that might predict surgical outcome; 2) whether a single random GH measurement is as useful a predictor as a mean GH value in determining disease activity pre- and postoperatively; 3) the interval time for the best postoperative assessment defined as the highest concordance rate between IGF-I levels and mean GH levels; and, thus, 4) the optimal biochemical strategy for defining the response to surgery.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All patients were assessed preoperatively in our center according to a consistent protocol (35). The diagnosis of acromegaly was based on the presence of relevant clinical features and confirmed by demonstration of failure of plasma GH to suppress to levels less than 1 mU/L (1 mU/L = 0.4 ng/mL) during an 75-g oral glucose tolerance test (OGTT; Refs. 27 and 36, 37, 38). GH secretion was assessed by calculating the mean serum GH from a five-point GH "day curve," samples being collected at 0830 h, 1100 h, 1300 h, 1700 h, and 1900 h, and basal serum IGF-I measurements. Non-GH-related pituitary function (serum T4, TSH, 0900 h cortisol, PRL, LH, FSH, testosterone, or estradiol) was assessed routinely pre- and postoperatively. The pituitary-adrenal axis was also assessed dynamically by examining the cortisol response to hypoglycemia during an insulin tolerance test, or during a glucagon test. All patients underwent preoperative imaging of the pituitary fossa with either computed tomography or magnetic resonance scanning. Pituitary adenomas were identified in all cases and were classified into microadenomas (intrasellar tumors of <1 cm diameter), mesoadenomas (tumors >1 cm diameter that do not touch the optic chiasm or invade the sphenoidal or cavernous sinuses), or macroadenomas (tumors >1 cm in diameter with suprasellar, inferior, or lateral extension compressing the surrounding structures).
TSS was performed by two surgeons (F.A. and I.S.), via a sublabial or transnasal approach, using an image intensifier. Obvious tumor was removed with attempted preservation of apparently normal pituitary tissue. After removal of the tumor, routine histopathological analysis was performed to establish the pattern of reticulin staining plus routine immunostaining for all pituitary hormones.
At 7 days postoperatively, serum GH was measured during a GH day curve, as described previously (34). Serum IGF-I levels were measured at least 1 week after surgery; 65% of patients had IGF-I measured 1–4 weeks after surgery (median, 7 days; 25th–75th percentiles, 6–10), the remaining 35% later than 4 weeks (median, 90 days, 30–180). When pituitary deficiency was recorded it was adequately replaced.
To assess whether a single random GH measurement provides as accurate an estimate of tissue exposure to GH as the mean GH value obtained from a GH day curve, we correlated each single basal serum GH level obtained from the five-point GH day curve, pre- and postoperatively. We also evaluated pre- and postoperatively the percentage distance of the minimum and the maximum of the set from the mean of the day, and the coefficient of variation in all patients. Furthermore, we analyzed the differences in these indices of fluctuation of GH levels in groups according to surgical outcome. Mean GH levels (based on a five-point day curve) were also obtained from 20 healthy volunteer subjects (age, 23.4 ± 1.8; body mass index, 22.9 ± 1.8; M/F = 12/8).
Assays
Serum GH was measured by immunoradiometric assay with reagents supplied by North East Thames Regional Radiommunoassay Laboratory (London, UK). The functional sensitivity is less than 0.5 mU/L with inter- and intra-assay coefficients of variation of less than 5%. The assay is monitored by the United Kingdom National External Quality Control Scheme. Serum IGF-I was measured using an in-house RIA after formic-acid acetone extraction, as described previously (39).
Immunostaining of pituitary hormones
Each pituitary tissue sample was stained with hematoxylin and eosin as a general stain to establish the various tissue components. A reticulin stain was performed to provide an indication of tissue architecture. Pituitary hormone stains were performed to establish the hormone phenotype of the tumor samples, as described previously (40, 41).
Statistical analysis
All data are presented as median values accompanied by 25th and 75th percentiles, or else as mean ± SE. The daily fluctuation of serum GH levels was assessed pre- and postoperatively by calculating the percentage distance of peak and nadir from the mean of a five-point day curve (2-hourly sampling) and the coefficient of variation. Nonparametric tests were used in comparing groups: the Mann-Whitney U test and its generalization by Kruskal and Wallis for comparisons between independent groups; the Wilcoxon matched-pairs signed-rank sum test was used to compare within subjects data before and after TSS (42). The correlation between variables was described using Spearman’s rank correlation coefficients. Logistic regression was used to analyze the relationship between cure after TSS (based on mean serum GH) and clinical variables. To analyze the overall influence of predictors of persistence, a multivariate logistic regression analysis was performed (43). Results were considered statistically significant if the two-tailed P was less than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In 19 patients this was by computed tomography, and in 48 patients
by magnetic resonance imaging scanning of the pituitary fossa.
Macroadenomas were found in 42 patients (63%), and mesoadenomas in 8
patients (12%); patients with macroadenomas had significantly higher
levels of GH compared with the 17 patients (25%) with microadenomas
(Table 1
). In patients with macroadenomas
the mean age at presentation was significantly lower than in patients
with microadenomas (41.0 ± 14.3 vs. 49.8 ± 10.9
yr, P < 0.05). There was no difference in gender, body
mass index, or apparent duration of disease among patients with
differently sized pituitary tumors.
|
There was an overall significant correlation between each of the
five single random GH measurements obtained from the GH day curve and
the mean GH value (r = +0.93 to 0.97, P <
0.0001), as well as between the maximum and the minimum value (r =
+0.94 to 0.98, P < 0.0001). However, within subjects
there was a clear fluctuation of the single GH values around the mean,
with a mean percentage distance from the maximum of 33 ± 3%
(range, 3–131) and from the minimum of 30 ± 2% (range, 4–81;
Fig. 1A). Preoperatively, no
statistically significant difference in the percentage distance from
the mean was found when patients were grouped according to the outcome
of surgery. There was a weak but significant correlation between log
mean GH and IGF-I levels (r = +0.48, P < 0.01;
Fig. 2), but there were large
discrepancies in individual values in many patients. Thirty-seven
patients in total were found to have at least one pituitary hormonal
deficiency preoperatively. Ten patients had partial hypopituitarism
(gonadotropin and ACTH deficiency), and one patient had
panhypopituitarism; another 22 patients had gonadotropin deficiency
whereas 4 patients had ACTH deficiency alone. Twenty-nine patients with
pituitary deficiency had macroadenomas, four mesoadenomas, and four
microadenomas on pituitary imaging.
|
|
We have subdivided our patients’ "response" to surgery
according to the mean postoperative GH levels. The preoperative levels
of mean GH (median, 71.4; 25th–75th percentiles, 26.4–113.3), IGF-I
(829, 599–1024), and PRL (275, 174–710) were significantly lowered by
TSS surgery with an 88% reduction in the mean GH (median, 8.8;
25th–75th percentiles, 4.6–18; P < 0.01), a 52%
reduction in IGF-I (400, 265–643; P < 0.01), and a
50% reduction in PRL levels (139, 92–207; P < 0.01).
Table 2 shows the percentages of patients
who achieved a postoperative mean serum GH level of less than 10 mU/L
and less than 5 mU/L, respectively. In this study, 23 patients (34%)
achieved the safe postoperative mean GH level of less than 5 mU/L or
less than 2 ng/mL (median, 2.2 mU/L; 25th–75th percentiles, 1.4–4.7
mU/L); thirteen additional patients (19%) had a reduction of serum GH
levels to less than 10 mU/L; thirty-one patients (46%), while
achieving a substantial reduction from preoperative mean GH levels
[median, 92.6 mU/L (41.7–150) to 20.6 (14.1–42.4)], had a
postoperative mean GH higher than 10 mU/L. After surgery, 20 of the 23
patients who had achieved a postoperative GH less than 5 mU/L obtained
a serum IGF-I value within the normal range adjusted for age (median,
251; range, 196–358 ng/mL). In addition, there were eight patients
with postoperative mean GH levels more than 5 mU/L (median, 15.6 mU/L)
who also obtained postoperative IGF-I levels within the normal range
(median, 266 ng/mL, 210–301). Overall, 28 patients (42%) obtained a
mean postoperative IGF-I level within the age-related normal range
(median, 245 ng/mL; 25th–75th percentiles, 208–295; Fig. 3). Figure 4 shows discrepancies between the two
criteria for remission of disease, mean serum GH levels, and serum
IGF-I, according to how long after TSS serum IGF-I was measured. The
time of assessment did not significantly affect the remission rate
(Fig. 4). Overall, 87% of patients with a mean GH less than 5 mU/L had
a normal IGF-I, whereas, respectively, 85% and 81% of patients with
mean GH 5–10 mU/L and more than 10 mU/L had high IGF-I levels,
indicating that setting the remission cut-off level at 5 mU/L (2 ng/mL)
has its biological correlate in the normalization of IGF-I levels. In
addition, the time of measurement of IGF-I seems not to be responsible
for the discrepancies between the two criteria. There was a significant
correlation between postoperative mean GH and postoperative IGF-I
levels (r = +0.68, P < 0.001) for the group as a
whole (Fig. 2
). The relationship was linear when serum GH was expressed
in a logarithmic scale.
|
|
|
Sixty-six of the 67 patients were found to have a pituitary tumor.
In the single patient without obvious tumor plasma GHRH levels were
normal, but GH levels did not improve after surgery. An area of
lymphocytic hypophysitis and a Rathke’s pouch cyst were found in
another two patients. In 43% of the patients immunostaining of their
tumors was positive for both GH and PRL; in 28% for GH alone; in 10%
for GH, PRL, and 
-subunit; in 9% for GH and -subunit; whereas in
3% it was positive for GH, PRL, -subunit, and ß-hCG. Tumors that
immunostained positive for PRL were associated with higher serum PRL
levels compared with those that were negative [531 mU/L (240–898)
vs. 189 mU/L (157–321), P < 0.01].
However, there was no difference in PRL levels between patients who had
micro- and macroadenomas [218 mU/L (161–652) vs. 321 mU/L
(190–712), P = not significant].
Factors predicting outcome
The outcome for different tumor sizes is shown in Table 2.
Preoperative tumor size seemed a major determinant of the outcome of
surgery, as 10 of 17 (59%) patients who had microadenomas achieved
postoperative GH levels less than 5 mU/L, compared with 11 of 42 (27%)
patients with macroadenomas; also, 2 of 8 patients with mesoadenomas
achieved postoperative GH levels less than 5 mU/L. Because tumor size
correlated with pretreatment mean GH levels, patients who were rendered
safe (mean, GH <5 mU/L) had significantly lower mean preoperative GH
levels [median 31 mU/L (25th–75th percentiles, 10.6–92.9)
vs. 78.5 mU/L (42.1–121.37), P < 0.01]
compared with patients who were not. Furthermore, the size of the tumor
correlated with the degree of pituitary deficiency preoperatively, with
patients with larger tumors having more significant pituitary
impairment than patients with smaller tumors. Lateral extension and
cavernous sinus invasion were independent factors associated with a
worse surgical outcome, as four of five patients who showed no change
in postoperative GH level showed such involvement. A logistic
regression analysis showed that tumor size was a significant
determinant of outcome (b = +0.70;
SE(b) = 0.29; P =
0.017), whereas preoperative IGF-I levels, although weakly correlated
with tumor size (r = +0.36, P < 0.05 Table 1),
did not predict the surgical outcome (Table 2). Preoperative PRL levels
were higher in patients who failed to achieve a surgical cure [498
mU/L (187–857) vs. 196 mU/L (136–315), P
< 0.01]; furthermore, PRL levels correlated with the presence of
positive tumor immunostaining for PRL rather than the size of the
tumor. A multivariate logistic regression model was generated to
evaluate the overall predictive value of preoperative mean GH, PRL, and
tumor size on surgical outcome. Mean preoperative serum GH
(b = +2.16; SE(b)
= +0.77; P = 0.005) and PRL (b = +2.13;
SE(b) = +0.83; P
= 0.010) were significant independent predictors of the outcome,
whereas tumor size was no longer significant, mainly because of its
interaction with the variable of serum GH. Interestingly, the effect of
PRL on outcome was independent of tumor immunostaining, which was not a
significant predictor. The incidence of noncurative surgery was not
statistically related to study years.
Mean GH measurement in normal subjects vs. patients with acromegaly
Spontaneous GH secretion in normal subjects (based on the
five-point day curve) was characterized by discrete pulses separated by
at least one nadir GH concentration below the limit of detection of the
assay (<0.5 mU/L). By contrast, in all untreated patients with
acromegaly the nadir GH was above the limit of sensitivity of the
assay. Eight of the 23 patients with postoperative mean GH levels less
than 5 mU/L showed at least one undetectable GH value; all these eight
patients also showed normal IGF-I levels 5–180 days after surgery
(Fig. 1B).
Mean vs. single random GH measurement
There was a significant correlation between both pre- and
postoperative mean GH and each single random GH, minimum GH, and
maximum GH values (postoperative mean GH against maximum and minimum GH
values r = +0.97 and r = +0.96, P < 0.01
respectively, Fig. 5). However, there
were 9 of 23 patients (39%) who obtained a mean postoperative GH value
less than 5 mU/L who had at least one single random GH value more than
5 mU/L; in addition, there were also 8 of 44 patients (18%) with a
mean postoperative GH value more than 5 mU/L who had at least one
single random GH value less than 5 mU/L (Table 3). Positive and negative predictive
values for at least one of the random GH levels being above or below 5
mU/L during a day curve were, respectively, 80% and 63% when disease
was assessed using the mean GH less than 5 mU/L, but 68% and 61% when
disease was assessed by normalization of IGF-I levels. The sensitivity
and specificity of mean GH against normalization of IGF-I were 92% and
71%, respectively, whereas positive and negative predictive values
were 82% and 87%, respectively. Analysis of fluctuation of
postoperative random GH values around their mean showed that the
percentage distances of the peak (55 ± 6%) and the nadir
(37 ± 2%) from the mean GH value were significantly higher than
they were preoperatively (P < 0.05; Fig. 1B).
Furthermore, postoperatively the percentage variation from the mean was
1.5- to 2-fold higher in the group of patients in remission (from
-46 ± 5% to 85 ± 11%) than it was before surgery, or in
patients with persistence of disease after TSS (from -32 ± 8%
to 39 ± 5%, respectively; P < 0.01).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Using this therapeutic aim, 34% of our patients, the majority of which had macroadenomas, achieved a safe mean GH level after TSS, and such patients will not require further adjunctive treatment. Whereas our study shows a very high correlation between single random GH levels and mean GH levels in patients with acromegaly, it also demonstrates the wide variation of single random GH measurement, and identifies potential mistakes that can arise when only single random GH measurements are used in evaluating patients after surgery. Nine of the 23 patients, who were rendered safe following TSS according to the above criterion, would have been erroneously diagnosed as partially treated if based on a single GH measurement alone and may have received further unnecessary treatment with radiotherapy and/or medication; in addition, 8 patients who did not achieve a postoperative mean GH level of less than 5 mU/L would also have been erroneously considered as safe if based on a single GH measurement alone and would have escaped further treatment; such patients would be exposed to the risk of long-term complications and mortality associated with mean GH levels more than 5 mU/L (5).
Based on the good correlation that exists between a single GH and mean GH values (14, 15, 20, 24, 44), several studies have previously questioned the necessity of mean GH values and have suggested reliance on single GH measurements in assessing disease activity in acromegaly (13, 15, 44). This was further reinforced when patients with acromegaly who were followed up with random single GH measurements were reported to have normal morbidity and mortality rates (18). However, secretion of GH is pulsatile, and only the mean of a series of GH values throughout the day has been shown to have highest sensitivity and specificity (24, 45); it has been reported that, based on single GH measurements alone, only 8 of 10 patients with active disease (high IGF-I) will be identified (24). Furthermore, in their long-term follow-up study based on a single random GH measurement, Abosch et al. (18) reported normal morbidity and mortality even in patients with GH values between 5 and 10 mU/L, a finding that differs from several other studies that have shown that only GH levels either less than 5 mU/L or less than 6 mU/L are associated with normal IGF-I secretion and mortality rates (5, 6, 7). Furthermore, in the study of Bates et al. (5), patients with mean GH levels more than 5 mU/L but less than 10 mU/L still showed increased morbidity and mortality rates. The contradiction in the study of Abosch et al. (18) was originally attributed to the fact that the majority of their patients harbored microadenomas, as assessed by older and less sensitive imaging techniques, and that the GH measurements were performed in a number of different laboratories and, thus, may not reflect current experience with sensitive imaging and assays (10). However, it is possible that the patients with single random GH levels less than 10 mU/L and normal mortality rates in long-term follow-up described by Abosch et al. (18) may well have had a mean GH value less than 5 mU/L; such a mean GH value would be compatible with normal survival irrespective of the fluctuation of random singleton GH values, as demonstrated in the patients who obtained a postoperative mean GH value less than 5 mU/L in our study (5). We have shown here that this fluctuation in GH values is, indeed, greater in those patients with safe mean GH levels.
Singleton random GH measurements also do not provide information on GH secretory dynamics. Several authors consider the post-OGTT nadir GH value as the best indicator of disease activity in acromegaly, because it also evaluates GH secretory dynamics (10, 13, 14). Although post-OGTT nadir GH values were not routinely available in our study, there is a good correlation of both mean GH and post-OGTT nadir GH value with the IGF-I level (26). However, a potential limitation on relying solely on post-OGTT nadir GH levels is that the majority of information on morbidity and mortality rates in acromegaly has been obtained from studies that used either mean GH (5) or single random GH measurements to assess disease activity (6, 18); there is relatively little information based on post-OGTT nadir GH levels (7). A normal postoperative OGTT may not be predictive of long-term cure, in particular because patients with a "normal" OGTT and subsequent clinical recurrences have been described (46, 47, 48). Indeed, recent data showed that nadir GH levels after glucose suppression measured by sensitive assays in patients with persistent acromegaly are much lower than previously recognized (27). In addition, the mean GH level may also be able to provide information on GH dynamics, because only patients with acromegaly who were cured obtained occasionally undetectable (<0.5 mU/L) GH levels and, thus, showed restored GH secretory dynamics similar to that seen in normal subjects.
Previous studies (19, 24, 45, 46) have demonstrated the pulsatile secretion of GH in patients with acromegaly. This is of particular interest because the increase in integrated GH concentration is accounted for mainly by an increase in the nonpulsatile fraction that most strongly correlates with serum IGF-I concentration (24, 45); twenty-four-hour sampling studies showed that in patients with acromegaly GH levels fail to fall to undetectable levels and the fluctuation of GH levels is usually around an increased baseline (24). Therefore, it is unlikely that a single GH measurement pretreatment will misclassify patients unless basal secretion is low. However, following successful treatment, patients with acromegaly may restore normal GH dynamics with a reduction of basal GH secretion, and a single GH may not reflect disease activity as shown from the results of our study. It seems reasonable to suggest that when assessing patients with active acromegaly a single random GH may suffice to reflect overall GH secretion, whereas when defining response to treatment and cure in patients with a substantial reduction of elevated GH levels, then a mean GH value has the appropriate sensitivity and specificity to reduce the number of false negative and/or false positive observations.
IGF-I, which has a long half-life and is GH dependent, has been proposed as a useful marker of overall GH status and, hence, disease activity. Although initial studies showed that GH was not correlated with IGF-I, recent evidence (19, 24, 26, 46) has suggested that IGF-I measurements could be used to monitor disease activity overall and predict long-term outcome in acromegaly (7, 11). Using normalization of IGF-I levels following surgery as a definition of cure, 42% of our patients were cured as opposed to 34% when based on a mean GH curve less than 5 mU/L, a figure closer to most cure rates currently quoted (13, 14, 49). However, there are well documented discrepancies between mean GH and IGF-I levels in individual patients (15, 26, 32), even in those with postoperative GH levels less than 0.5 mU/L (24, 50). Until the significance of these findings with respect to long-term outcome is known, we suggest that estimates of serum GH should continue to be used in the follow-up of these patients in addition to IGF-I. Although there is a much more extensive body of data relating long-term prognosis in acromegaly to GH levels rather than IGF-I, current data suggest that both normalization of IGF-I levels and postoperative mean serum GH levels less than 5 mU/L are associated with normal morbidity and mortality rates; if both of these targets are achieved, then the evidence suggests that such patients will not require any immediate further treatment (7, 10, 20). The response to surgery was seen to depend principally on tumor size, serum GH, and PRL levels; as these are in turn correlated, surgical outcome can be related to serum GH and PRL levels alone. This is also true in relation to the presence of pituitary deficiencies. By contrast, we have found that basal preoperative serum IGF-I, which correlated poorly with tumor size, was unrelated to surgical outcome. Total IGF-I has a half-life a little less than 24 h, and we reasoned that measurement at a minimum of 1 week postoperatively should reflect surgical efficacy. Indeed, we found no difference in apparent cure rates assessed 1–3 weeks and more than 4 weeks after surgery, despite suggestion that such early assessment may be inaccurate (11). In a previous study, it was reported that early normalization of IGF-I was generally associated with persistent cure, although a gradual fall with time could also occur (30). Until data are available on longitudinal changes in time following surgery, we suggest that early assessment of postoperative IGF-I is prognostically useful.
In summary, we suggest that the mean GH day curve currently remains a valuable way of assessing disease activity in acromegaly following surgery because it eliminates the limitations of random GH measurements, may reflect GH secretory dynamics, predicts normal IGF-I levels in the majority of cases, and provides information regarding long-term morbidity and mortality rates. Although random single GH measurements show a high correlation with mean GH levels, individual variations may lead to misdiagnosis of the true level of GH. Estimation of both mean GH and IGF-I levels may be necessary to follow-up patients with discordant results to optimize management. Preoperatively, the size of the tumor, and degree of anterior pituitary deficiency, GH, and PRL levels, but not IGF-I levels, can be used to help predict surgical outcome and, thus, identify in advance patients who will be likely to require further treatment to achieve safe GH levels.
In practice, any postoperative mean GH more than 5 mU/L with an IGF-I level above the age-reference range clearly requires further treatment. When biochemical data are not congruent at an early assessment, we suggest reevaluation at 6 months, bearing in mind those conditions that may result in altered IGF-I levels. In our opinion, a mean serum GH more than 5 with IGF-I levels between the median and the upper limit of normal age reference range should be considered for medical treatment, whereas a mean GH more than 10 mU/L should be treated aggressively. In the presence of a mean GH less than 5 mU/L with elevated IGF-I levels, the decision is mainly clinical based on age and cardiovascular, metabolic, and tumoral comorbidity.
|
Footnotes |
|---|
Received September 13, 2000.
Revised December 18, 2000.
Accepted December 28, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  O. Alexopoulou, M. Bex, R. Abs, G. T'Sjoen, B. Velkeniers, and D. Maiter Divergence between Growth Hormone and Insulin-Like Growth Factor-I Concentrations in the Follow-Up of Acromegaly J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1324 - 1330. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Bajuk Studen and A. Barkan Assessment of the Magnitude of Growth Hormone Hypersecretion in Active Acromegaly: Reliability of Different Sampling Models J. Clin. Endocrinol. Metab., February 1, 2008; 93(2): 491 - 496. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Pokrajac, A. G Claridge, S K A. Shakoor, and P. J Trainer The octreotide test dose is not a reliable predictor of the subsequent response to somatostatin analogue therapy in patients with acromegaly Eur. J. Endocrinol., February 1, 2006; 154(2): 267 - 274. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Fahlbusch, B. v Keller, O. Ganslandt, J. Kreutzer, and C. Nimsky Transsphenoidal surgery in acromegaly investigated by intraoperative high-field magnetic resonance imaging Eur. J. Endocrinol., August 1, 2005; 153(2): 239 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Ayuk, R. N. Clayton, G. Holder, M. C. Sheppard, P. M. Stewart, and A. S. Bates Growth Hormone and Pituitary Radiotherapy, But Not Serum Insulin-Like Growth Factor-I Concentrations, Predict Excess Mortality in Patients with Acromegaly J. Clin. Endocrinol. Metab., April 1, 2004; 89(4): 1613 - 1617. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. M. Holdaway, R. C. Rajasoorya, and G. D. Gamble Factors Influencing Mortality in Acromegaly J. Clin. Endocrinol. Metab., February 1, 2004; 89(2): 667 - 674. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Swords, C. A. Allan, P. N. Plowman, A. Sibtain, J. Evanson, S. L. Chew, A. B. Grossman, G. M. Besser, and J. P. Monson Stereotactic Radiosurgery XVI: A Treatment for Previously Irradiated Pituitary Adenomas J. Clin. Endocrinol. Metab., November 1, 2003; 88(11): 5334 - 5340. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. De, D. A. Rees, N. Davies, R. John, J. Neal, R. G. Mills, J. Vafidis, J. S. Davies, and M. F. Scanlon Transsphenoidal Surgery for Acromegaly in Wales: Results Based on Stringent Criteria of Remission J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3567 - 3572. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z Merza Modern treatment of acromegaly Postgrad. Med. J., April 1, 2003; 79(930): 189 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Trainer Acromegaly--Consensus, What Consensus? J. Clin. Endocrinol. Metab., August 1, 2002; 87(8): 3534 - 3536. [Full Text] [PDF]
|
|
|
|
 |
|
 T.-J. Wu, E. H. Yu, L. Song, R. L. Taylor, and P. C. Kao Filter Paper Collection of Plasma for IGF-I Test in Patients with Acromegaly Ann. Clin. Lab. Sci., July 1, 2002; 32(3): 287 - 291. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||
) around their mean (
) in all
patients. The GH levels are plotted in the common logarithmic scale.
) or IGF
(
) postoperative
levels, with the total number of tumors (
) related to size. Micro,
Intrasellar microadenoma; Meso, mesoadenoma; Macro, macroadenoma.
