J Neurosurg 97:287 292, 2002 Changing patterns of insulin-like growth factor I and glucose-suppressed growth hormone levels after pituitary surgery in patients with acromegaly ANA LAURA ESPINOSA-DE-LOS-MONTEROS, M.D., MOISÉS MERCADO, M.D., ERNESTO SOSA, M.D., OSCAR LIZAMA, M.D., GERARDO GUINTO, M.D., BLAS LOPEZ-FELIX, M.D., OSCAR GARCIA, M.D., IRMA HERNÁNDEZ, M.D., ALFONSO OVALLE, M.D., AND VICTORIA MENDOZA, M.D. Endocrinology and Neurosurgery Sections, Hospital de Especialidades, Centro Medico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, México City, México A Object. According to a recent consensus statement on the treatment of acromegaly, its biochemical cure is defined as the normalization of age- and sex-adjusted insulin-like growth factor (IGF) I levels and the suppression of growth hormone (GH) by glucose to lower than 1 ng/ml. The present study was prompted by the clinical observation that many cases of acromegaly can be considered cured according to one criterion but not others at different moments in a patient s postoperative course. Methods. Fifty-three patients with acromegaly (30 women and 23 men) harboring nine microadenomas and 44 macroadenomas were evaluated after surgery by assessing age- and sex-adjusted IGF-I levels as well as glucose-suppressed GH levels. Fifty of these patients were studied more than once during follow up. Acromegaly was categorized as cured if the patient s IGF-I level was normal and their glucose-suppressed GH level was lower than 1 ng/ml; the disease was considered to be active if the patient s IGF-I level was high and the GH nadir was higher than 1 ng/ml following administration of glucose. Discordant categories of the disease were found in patients with high IGF-I levels and a GH nadir lower than 1 ng/ml after glucose administration and in those with normal IGF-I levels and a GH nadir higher than 1 ng/ml after glucose intake. At the first postoperative biochemical evaluation (1 3 months), 34% of patients harboring macroadenomas were classified as having been cured of acromegaly, 39% as having the active disease, and 27% as having the discordant form of the disease. When last evaluated ( 12 months postoperatively), the percentage of patients with the discordant form dropped to 14% and the proportion of cases cured and active was 44% and 41%, respectively. Of the nine patients with microadenomas, 44.4% were cured of acromegaly, 33.2% had the active disease, and 22% had the discordant variety on first evaluation. Twelve months or longer after transsphenoidal surgery, 55.5% of cases were cured, 11.1% were active, and 33% were discordant. In most cases, the discordant variety developed because of a persistently elevated level of IGF-I, followed by an incompletely suppressed GH level. Nineteen patients (38%) modified their biochemical category. In 15 of these patients this change in category was due to a change in IGF-I levels, becoming normal in 12 patients and rising to above normal range in three. A tumor remnant was demonstrated on magnetic resonance images in only four of these 19 patients. Conclusions. The authors conclude that the discordance rate between the biochemical markers that define cure in acromegaly is higher than previously reported, and the biochemical status assigned to a patient early in the postoperative course is very likely to change later, particularly when initially discordant. KEY WORDS insulin-like growth factor I growth hormone acromegaly transsphenoidal surgery outcome Abbreviations used in this paper: GH = growth hormone; IGF-I = insulin-like growth factor I. CROMEGALY is a rare disorder, with an estimated prevalence rate of approximately 40 cases per million. 3 In more than 95% of patients, this disease results from a GH-secreting pituitary adenoma. 10 The long-term clinical and biological consequences of excess GH include musculoskeletal deformities, acral enlargement, impaired carbohydrate metabolism, hypertension, and an increased mortality rate largely due to cardiovascular events and perhaps neoplastic causes. 12,14,15 Despite the development of somatostatin analogs, GH receptor antagonists, and sophisticated methods of pituitary irradiation, transsphenoidal surgery remains the treatment of choice. 11 Biochemical criteria for cure of acromegaly have become more stringent as our knowledge of the regulation of the somatotropic axis increases. Overall, the cure rate ranges from 70 to 80% for microadenomas and from 30 to 45% for macroadenomas. The biochemical parameters used to determine the cure of acromegaly have evolved from a simple decrease in postoperative GH levels, to the ability of a glucose load to suppress GH secretion, to the use of GH-dependent peptides such as IGF-I and IGF binding protein- 3. 13,17 Researchers at some neuroendocrinological centers consider the normalization of IGF-I levels as the only crite- J. Neurosurg. / Volume 97 / August, 2002 287
A. L. Espinosa-de-los-Monteros, et al. FIG. 1. Early (1 3 months), intermediate (4 11 months), and late ( 12 months) postoperative biochemical evaluations in patients with microadenomas. Upper: Graphs depicting the actual GH nadir (GHn) after administration of 75 g of glucose plotted against the patient s IGF-I level divided by the upper limit of normal age- and sex-adjusted values. Lower: Tables showing the percentage of cases cured, active, and discordant at each evaluation. The actual number of patients in each category is shown in parentheses. N = number of patients. rion for biochemical cure, whereas others use different levels of glucose-suppressed GH values as the gold standard. 4 Due to the lack of long-term studies in which strict biochemical criteria are used, there is no agreement regarding the most appropriate time to determine the cure of acromegaly after transsphenoidal surgery. Answers to questions such as how long it takes the somatotropic axis to return to normal (that is, normalization of IGF-I levels and complete GH suppression by glucose) after successful surgical treatment of acromegaly remain elusive. The present study was prompted by the lack of uniform criteria to define the cure of acromegaly and by the clinical observation that many cases can be considered cured according to one biochemical parameter but not by others at different moments during the patient s postoperative course. Clinical Material and Methods Patient Population and Study Design A group of 53 patients, consisting of 30 women and 23 men whose ages ranged from 22 to 80 years (mean 42.5 12.3 years), with newly diagnosed acromegaly treated using transsphenoidal surgery between January 1997 and June 1999, were included in this study. Transsphenoidal surgery was performed in patients harboring macroadenomas via the sublabial approach, whereas in patients with microadenomas the transnasal route was chosen. Magnetic resonance images obtained in most patients (44) demonstrated tumors larger than 1 cm, but only one had evidence of chiasm compression and cavernous sinus invasion. In all cases, the GHsynthesizing nature of the tumor was confirmed using immunohistochemistry. Biochemical evaluations took place at 8:00 a.m., after patients completed an overnight fast. An indwelling catheter was inserted into an antecubital vein and blood was drawn to determine baseline levels of IGF-I and GH as well as GH levels at 30, 60, 90, and 120 minutes after the patient drank 75 g of glucose. All patients remained seated for the entire procedure. Three patients underwent only one postoperative biochemical evaluation. Hormone Assays Growth hormone was measured using a solid-phase, twosite chemiluminescent enzyme immunometric assay with the aid of an automated analyzer (IMMULITE; Diagnostic Products Corp., Los Angeles, CA) having a detection limit of 0.01 ng/ml and intra- and interassay coefficients of variation of 6%. Insulin-like growth factor I was separated 288 J. Neurosurg. / Volume 97 / August, 2002
Hormone levels after pituitary surgery in patients with acromegaly from its binding proteins by using an acid-ethanol extraction procedure, and IGF-I levels were then quantified in the extracted samples by performing a two-site immunoradiometric assay (Diagnostic Systems Laboratory, Webster, TX) with intra- and interassay coefficients of variation of 2.6% and 4.4%, respectively. Normal age- and sex-adjusted reference values used in our laboratory for IGF-I are as follows: 20 to 25 years, 215 to 590 ng/ml in men and 110 to 521 ng/ ml in women; 26 to 30 years, 120 to 480 ng/ml in men and 129 to 502 ng/ml in women; 31 to 40 years, 100 to 470 ng/ ml in men and 130 to 354 ng/ml in women; 41 to 50 years, 100 to 300 ng/ml in men and women; and 51 to 70 years, 78 to 260 ng/ml in men and women. Definitions of Biochemical Outcomes Acromegaly was considered cured if the GH nadir was below 1 ng/ml after glucose administration and the age- and sex-adjusted IGF-I level was normal. Patients who had postoperative GH nadirs above 1 ng/ml after glucose intake and elevated age- and sex-adjusted IGF-I levels were defined as having the active form of the disease. Two discordant categories of acromegaly were established: a GH nadir lower than 1 ng/ml after glucose intake combined with an elevated IGF-I level, and a GH nadir greater than 1 ng/ ml after glucose administration together with a normal IGF- I level. With respect to time, an early evaluation was defined as that performed 1 to 3 months after transsphenoidal surgery, intermediate when performed 4 to 11 months after surgery, and late if it was 12 months or longer after surgery. Statistical Analysis Descriptive analysis of the data was expressed in proportions, and a bivariate analysis between IGF-I and glucosesuppressed GH levels was performed using a concordant phi coefficient. Results Three patients underwent only one postoperative biochemical evaluation; two were considered to have concordantly active acromegaly at 9 and 12 months, and the disease was considered to be concordantly cured at 24 months in one patient. The remaining 50 patients, 41 of whom harbored macroadenomas and nine of whom had microadenomas, underwent at least two biochemical evaluations. At the initial evaluation, acromegaly was cured in a total of 36% of the patients, active in 38%, and discordant in 26% (an elevated age- and sex-adjusted IGF-I level combined with a GH nadir lower than 1 ng/ml after glucose administration). On the last biochemical evaluation, the disease was cured in 46% of patients, active in 36%, discordant with an elevated IGF-I level in 10%, and discordant with a GH nadir higher than 1 ng/ml after glucose intake in 8%. Of the nine patients with microadenomas, 44.4% of cases were considered to be cured, 33.3% were active, and 22.2% were discordant with an elevated IGF-I level on early postoperative biochemical evaluation (Fig. 1). Three of these nine patients had an intermediate evaluation within 4 to 11 months of transsphenoidal surgery; two cases were considered to be cured based on biochemical studies and one was J. Neurosurg. / Volume 97 / August, 2002 active. At the last biochemical evaluation, 55.5% of the nine cases of microadenomas were classified as cured, 11.1% as active, 22.2% as discordant due to a high IGF-I level combined with a GH nadir lower than 1 ng/ml after glucose intake, and 11.1% as discordant due to a normal IGF-I level together with a GH nadir higher than 1 ng/ml after glucose administration. During the early postoperative biochemical evaluation of 41 patients with macroadenomas, 34% of cases were considered to be cured, 39% active, and 26.8% discordant due to an elevated IGF-I level with a normal glucose-suppressed GH value (Fig. 2). Twenty-nine of these 41 patients underwent an intermediate evaluation, which revealed that 38% of cases were cured, 34% active, and 27.6% discordant due to a high IGF-I level together with a glucose-suppressed GH level lower than 1 ng/ml. At the last biochemical evaluation, 44% of macroadenomas were cured, 41.4% active, and 14.6% discordant (7.3% with an elevated IGF- I level and 7.3% with a glucose-suppressed GH level 1 ng/ml). Of the 50 patients who underwent more than one postoperative biochemical assessment, 31 (62%) remained in the same category throughout the study period, that is, 15 cured (12 macroadenomas, three microadenomas), 15 active (14 macroadenomas, one microadenoma), and one discordant (one macroadenoma with a GH nadir 1 ng/ml after glucose administration and a high IGF-I level). Nineteen patients harboring 14 macroadenomas and five microadenomas modified their biochemical category at a second, third, or fourth postoperative evaluation (Table 1). All 19 patients underwent postoperative follow up for longer than 1 year. In 15 of these patients, the change in category was due to a change in the level of IGF-I; that is, in 12 patients (nine with macroadenoma and three with microadenoma) the IGF-I level became normal, and in three patients (two with macroadenoma and one with microadenoma) the IGF-I level increased to values above the normal age- and sex-adjusted range. Eight patients (six with macroadenoma and two with microadenoma), who were initially considered to have the discordant variety of the disease because of a high IGF-I level together with a glucose-suppressed GH level lower than 1 ng/ml, eventually normalized their IGF-I level. The disease in these patients was reclassified as cured at their last follow up, which was completed at a mean of 11.6 months after transsphenoidal surgery. In none of these eight patients was there evidence of a tumor remnant on imaging studies and all had some degree of clinical improvement of their disease (acral enlargement, headache, hypertension, or glucose intolerance). Conversely, three patients (two with macroadenoma and one with microadenoma) in whom the disease was considered to be cured at the first biochemical evaluation were reassessed with the discordant form of the disease because their IGF- I level rose to above normal range at their last follow up. None of them had a tumor remnant, and they all experienced some degree of clinical improvement of acromegaly. The GH nadir following glucose administration changed in five patients (four harboring macroadenoma and one with microadenoma) on further biochemical evaluation, becoming normal in only one (Table 1). Among the 19 patients who experienced a change in their postoperative biochemical category, only four demonstrated a tumor remnant; two had the active disease at their last follow up, and 289
A. L. Espinosa-de-los-Monteros, et al. FIG. 2. Early (1 3 months), intermediate (4 11 months), and late ( 12 months) postoperative biochemical evaluations in patients with macroadenomas. Upper: Graphs depicting the actual GH nadir after administration of 75-g glucose load plotted against the patient s IGF-I level divided by the upper limit of normal age- and sex-adjusted values. Lower: Tables showing the percentage of cases cured, active, and discordant at each evaluation. The actual number of patients in each category is shown in parentheses. the other two had the discordant form with normal IGF-I levels and GH nadirs higher than 1 ng/ml following administration of glucose. Discussion According to a recently published consensus statement on the treatment of acromegaly, biochemical cure is defined as a restoration of disordered GH secretion to normal levels. 7 This consensus clearly states that cure implies both a normalization of the age- and sex-adjusted IGF-I level and a GH nadir lower than 1 ng/ml after an oral glucose load. Two items that remain unclear, however, are how often these two biochemical goals can be achieved simultaneously in one patient and the optimal time to evaluate outcome. Most of the available information regarding biochemical cure in acromegaly comes from retrospective studies that use different criteria to define remission of the disease. Our study demonstrates that the biochemical indices that define remission of acromegaly can be discordant in as many as 20% of patients. This discrepancy is the result of either a persistently elevated IGF-I level or an incompletely suppressed GH secretion and tends to be more evident when patients are evaluated early in the postoperative period. Although the issue of concordance between the IGF-I level and the glucose-suppressed GH level had not been previously evaluated in a systematic manner, some information can be obtained from recently published studies in which biochemical outcome has been assessed after transsphenoidal surgery in cases of acromegaly. Abosch, et al., 1 report long-term postoperative follow-up data in 101 patients, 13% of whom had an elevated IGF-I level even though their postoperative basal GH concentrations were below 2.5 ng/ml and the glucose-suppressed GH value was below 2 ng/ml. Similarly, Freda, et al., 6 report a discordance rate of approximately 12% between postoperative IGF-I and glucose-suppressed GH levels to lower than 2 ng/ml. The discrepancy in the results of these two studies and our own regarding the rate of discordance is very likely due to the use of more stringent GH suppression criteria ( 1 ng/ ml as opposed to 2 ng/ml) in our series. Data from a recent longitudinal follow-up evaluation of 75 patients with acromegaly in which a sensitive GH immunoradiometric assay was used support our findings. 5 In this study, almost one third of the patients who underwent surgery for acromegaly and who achieved a normal age- and sex-adjusted IGF-I level had a GH nadir following glucose intake significantly higher than that found in matched controls. Perhaps the most interesting of our findings is the change in biochemical category registered in the postoperative follow-up period in more than a third of the patients. In most 290 J. Neurosurg. / Volume 97 / August, 2002
Hormone levels after pituitary surgery in patients with acromegaly of these cases the category change was due to changes in the IGF-I level. In fact, of the 19 patients whose biochemical category was changed, 12 were due to a late normalization of IGF-I. Furthermore, eight of these 12 patients persisted with high age- and sex-adjusted IGF-I levels for several months after surgery, despite a progressive improvement of symptoms, a complete glucose-induced GH suppression, and the absence of tumor remnant on imaging studies. Although it is well known that IGF-I levels may remain elevated for several weeks after successful removal of a GH-secreting pituitary adenoma, 9 the magnitude of the delay that we observed in many of our patients had never been formally reported. In this regard, Kao, et al., suggested that IGF-I levels are useful in the postoperative evaluation of acromegaly when assessed 4 days after transsphenoidal surgery. This delay in the normalization of IGF-I after correction of the hypersomatotropinemia could be due to the persistence for variable periods of time of the altered metabolic environment that characterizes acromegaly (insulin resistance, hyperinsulinemia, impaired carbohydrate and lipid metabolism, and sex hormone levels). On a more speculative level, perhaps the GH receptor in target tissues acquires some degree of autonomy that results in GH-independent IGF-I synthesis. To our knowledge, there is no experimental proof of this, and the details of the GH/GHreceptor interaction after correction of longstanding hypersomatotropinemia have not been studied. The ability to suppress GH secretion after a glucose load also changed over time in a significant number of patients following transsphenoidal surgery. Although in several cases the degree of GH suppression improved, only one patient was eventually able to completely suppress GH to a level below 1 ng/ml. On the other hand, the degree of GH suppression worsened in four patients. Thus, in contrast to postoperative IGF-I levels, which tended to improve when they changed, the GH suppression by glucose tended to worsen over time. This observation seriously challenges the use of the glucose-suppressed GH level as the sole criterion for remission in acromegaly. We found a somewhat different biochemical behavior in patients harboring macroadenomas compared with those who harbored microadenomas. Surprisingly, in this group the discordance rate between IGF-I and glucose-suppressed GH levels increased at the last evaluation. The small number of patients with microadenomas in our series prevents us from drawing conclusions in this regard, however. We believe that our findings have several biological and clinical implications. If biological cure of acromegaly means restoring to normal the altered somatotropic axis, this goal is probably very hard to achieve. On the other hand, the timing of the evaluation is crucial in determining whether biochemical cure has been achieved. Our results show how dynamic the postoperative biochemical indices of acromegaly can be and prompt us to reconsider previous recommendations regarding the optimal postoperative evaluation time. 7 Growth hormone pulsatility study results demonstrating that patients cured of acromegaly, even those with normal IGF-I levels, do not recover a normal GH secretory pattern support this view. 8,16 From a clinical and epidemiological point of view, lowering the basal postoperative GH level to below 2.5 ng/ml reduces the increased mortality rate in these patients compared with that found in the general population. 2,15 Nevertheless, the persistence of a J. Neurosurg. / Volume 97 / August, 2002 TABLE 1 Change in postoperative biochemical category in 19 patients harboring 14 macroadenomas and five microadenomas* Disease Category Dis to Cured Active Dis to Dis to Factor Cured to Dis to Dis Active Dis total no. of pa- 8 (42) 3 (15.7) 4 (21) 3 (15.7) 1 (5.2) tients (%) no. of patients w/ microadenoma 2 1 2 0 0 no. of patients w/ macroadenoma 6 2 2 3 1 tumor remnant 0 0 1 2 1 disease category Y C C Y A X (3) Y A Y X change A Y (1) detection time in mos (range) 11.6 5.7 12 5.29 16.3 8 20 6.9 15 (5 21) (6 16) (8 24) (12 24) *A= active; C = cured; dis = discordant; X = GH nadir higher than 1 ng/ ml together with a normal IGF-I level; Y = GH nadir lower than 1 ng/ml with a high IGF-I level. Two macroadenomas, one microadenoma. One microadenoma. Values are expressed as the means standard deviation. biologically abnormal somatotropic axis may represent a constant risk for recurrence of the disease. Many of these patients with low-grade acromegaly have no tumor remnant demonstrated on imaging studies, and therefore a second transsphenoidal surgery or even radiotherapy cannot be offered for treatment. What is clear is that these patients, even those who are cured according to the strictest criteria available, should be followed up for their entire life. Whether medical interventions such as long-acting somatostatin analogs or GH-receptor antagonists will prevent recurrences in clinically improved but not biologically cured patients with acromegaly remains to be evaluated. References 1. Abosch A, Tyrrell JB, Lamborn KR, et al: Transsphenoidal microsurgery for growth hormone-secreting pituitary adenomas: initial outcome and long-term results. J Clin Endocrinol Metab 83:3411 3418, 1998 2. Bates AS, Van t Hoff W, Jones JM, et al: An audit of outcome of treatment of acromegaly. Q J Med 86:293 299, 1993 3. Bengtsson BA, Eden S, Ernest I, et al: Epidemiology and longterm survival in acromegaly. A study of 166 cases diagnosed between 1955 to 1984. Acta Med Scand 223:327 335, 1988 4. Clayton RN: New developments in the management of acromegaly. Should we achieve absolute biochemical cure? J Endocrinol 155 (Suppl 1):S23 S32, 1997 5. Freda PU, Landman RE, Post KD, et al: Longitudinal follow up evaluation of GH levels after OGTT with a sensitive IRMA in 75 postoperative acromegalic patients, in Program & Abstracts: The Endocrine Society s 82nd Annual Meeting June 21 24, 2000. Toronto, ON: The Endocrine Society, 2000, p 527 (Abstract) 6. Freda PU, Wardlaw SL, Post KD: Long-term endocrinological follow-up evaluation in 115 patients who underwent transsphenoidal surgery for acromegaly. J Neurosurg 89:353 358, 1998 7. Giustina A, Barkan A, Casanueva FF, et al: Criteria for cure of 291
A. L. Espinosa-de-los-Monteros, et al. acromegaly: a consensus statement. J Clin Endocrinol Metab 85:526 529, 2000 8. Ho PJ, Jaffe CA, Friberg RD, et al: Persistence of rapid growth hormone (GH) pulsatility after successful removal of GHproducing pituitary tumors. J Clin Endocrinol Metab 78: 1403 1410, 1994 9. Kao PC, Laws ER Jr, Zimerman D: Somatomedin C/Insulin like growth factor I levels after treatment of acromegaly. Ann Clin Lab Sci 22:95 99, 1992 10. Melmed S, Ho K, Klibanski A, et al: Clinical review 75: recent advances in pathogenesis, diagnosis, and management of acromegaly. J Clin Endocrinol Metab 80:3395 3402, 1995 11. Melmed S, Jackson I, Kleinberg D, et al: Current treatment guidelines for acromegaly. J Clin Endocrinol Metab 83:2646 2652, 1998 12. Molitch ME: Clinical manifestations of acromegaly. Endocrinol Metab Clin North Am 21:597 614, 1992 13. Ochoa R, Mercado M, Chacón X, et al: Usefulness of insulin-like growth factor binding protein-3 levels to determine acromegaly activity and effectiveness of transsphenoidal pituitary surgery. Arch Med Res 30:303 306, 1999 14. Orme SM, McNally RJQ, Cartwright RA, et al: Mortality and cancer incidence in acromegaly: a retrospective cohort study. United Kingdom Acromegaly Study Group. J Clin Endocrinol Metab 83:2730 2734, 1998 15. Rajasoorya C, Holdaway IM, Wrightson P, et al: Determinants of clinical outcome and survival in acromegaly. Clin Endocrinol 41:95 102, 1994 16. Semer M, Faria ACS, Nery M, et al: Growth hormone pulsatility in active and cured acromegalic subjects. J Clin Endocrinol Metab 80:3767 3770, 1995 17. Wass JAH: Growth hormone, insulin-like growth factor-i and its binding proteins in the follow-up of acromegaly. J Endocrinol 155 (Suppl 1):S17 S19, 1997 Manuscript received March 30, 2001. Accepted in final form April 22, 2002. Financial support was provided by the National Council for Science and Technology (CONACyT) Scholarship Program to Drs. Sosa and Ovalle. The results of this study were presented in part at the 82nd Meeting of the Endocrine Society, Toronto, Ontario, Canada, in June 2000. Address reprint requests to: Moisés Mercado, M.D., Endocrinology Section, Hospital de Especialidades, CMN S.XXI, Aristoteles 68, Polanco 11560, Mexico City, Mexico. email: mmercadoa@ yahoo.com. 292 J. Neurosurg. / Volume 97 / August, 2002