Revisiting the Role of Insulin-like Growth Factor-1 Measurement After Surgical Treatment of Acromegaly

Abstract

Context

In the management of growth hormone (GH)–secreting pituitary adenomas, the oral glucose tolerance test (OGTT) has been the gold standard not only for diagnoses but also for the determination of biochemical remission. Insulin-like growth factor-1 (IGF-1) is an essential biomarker, although it should be adjusted for both age and sex.

Objective

We evaluated whether IGF-1 levels could serve as a reliable alternative to an OGTT for disease monitoring after the surgical treatment of acromegaly. We retrospectively reviewed the medical records of 320 patients who underwent surgical resection of their GH-secreting pituitary tumors at the Severance hospital. Receiver operator characteristic (ROC) analyses were performed to validate the accuracy of IGF-1 levels for the assessment of remission. In addition, regression analyses were performed to identify factors associated with discrepancy between OGTT and IGF-1 levels.

Results

Except for 1 week after surgery, ROC analyses showed an area under the curve of greater than 0.8 for IGF-1 at all time points. Of 320 patients, 270 achieved endocrine remission after surgery alone. Among these patients, IGF-1 levels were normalized in 250 patients. The mean duration from surgery to IGF-1 normalization was 4.7 months. Regression analyses demonstrated that risk of failed IGF-1 normalization was increased by 3.1-fold when the tumor invaded the cavernous sinus and increased by 9.0-fold in patients with incomplete tumor removal.

Conclusion

IGF-1 level is a reliable alternative to OGTT and plays a valuable role in monitoring acromegaly status.

Acromegaly is mostly caused by growth hormone (GH)–secreting pituitary adenomas and characterized by elevated GH and insulin-like growth factor-1 (IGF-1) levels. Metabolic, cardiovascular, and musculoskeletal comorbidities frequently occur due to high levels of GH and IGF-1 in patients with acromegaly, and, for this reason, patients without adequate management have high mortality rates (1-3).

For GH-secreting pituitary tumors, surgical resection is recommended as the first-line treatment (4, 5). After surgical treatment, GH and IGF-1 levels are used to monitor whether biochemical remission has been achieved (2, 6, 7). Since the ultimate treatment goal for patients with acromegaly is to lower mortality by reducing cardiovascular, cerebrovascular, and respiratory complications, it would be ideal to monitor IGF-1 levels, which elicit more direct effects on complications than GH (8, 9). However, monitoring the surgical outcomes of GH-secreting pituitary tumors using IGF-1 levels is often unreliable. First, normal levels of IGF-1 vary with race, age, and sex, which make the interpretation of IGF-1 levels inconvenient (10, 11). Second, there is a delay in the stabilization of IGF-1 levels compared with those of GH; thus IGF-1-based assessments of remission are unreliable immediately after surgery (12-18). Third, IGF-1 measurements are an indirect way to monitor endocrine activity, as IGF-1 is synthesized in the liver after GH stimulation (19). For these reasons, the oral glucose tolerance test (OGTT) has been preferred as the gold standard in studies of the management of acromegaly in our institution.

Nevertheless, unlike GH, which undergoes pulsatile secretion, IGF-1 is a stable marker, which is its greatest advantage as a biomarker for monitoring. In this study, after the surgical removal of GH-secreting pituitary tumors, we evaluated whether endocrine remission could be assessed and monitored by the measurement of IGF-1 levels alone as a possible replacement for OGTT. In addition, we aimed to identify factors associated with discrepancies between IGF-1 levels and OGTT on the interpretation of remission after the surgical removal of GH-secreting pituitary tumors.

Materials and Methods

Study Population

This study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board at our institution. Written consent of the patients was waived due to the study’s retrospective nature. The medical records of patients who underwent surgical resection of their GH-secreting pituitary tumors between January 2011 and December 2018 at our institution were reviewed retrospectively. Patients with recurrent tumors, those younger than 20 years of age, and those with fewer than 3 postoperative OGTT and IGF-1 measurements were excluded from the study. A total of 320 patients who were surgically treated for their GH-secreting pituitary tumors at our institution were included in this study. Acromegaly was diagnosed based on biochemical findings (nadir GH level ≤1 ng/mL during a 75-g OGTT and elevated serum IGF-1 levels). An age- and sex-matched normal range of IGF-1 was obtained from a previous study on IGF-1 in healthy Korean adults (20).

All patients were evaluated using sellar magnetic resonance imaging (MRI) before and after surgery. After surgical resection, a histological examination was performed to confirm the presence of GH-secreting pituitary adenoma cells. All patients underwent transsphenoidal surgery (TSS) of their tumors, which was performed by 3 experienced surgeons. Pituitary adenomas were classified based on preoperative MRI and intraoperative findings, according to modified Hardy’s radiological classifications, as described previously (21, 22). In brief, a type I adenoma is a tumor confined entirely within the sella that is smaller than 1 cm in diameter. A type II adenoma is a tumor that has extended into the suprasellar area but less than 1 cm above the line of the diaphragm. A type III adenoma is a tumor that has extended into the suprasellar area more than 1 cm from the diaphragmatic border or into the sphenoid sinus. A type IV adenoma is a tumor that has clearly invaded either the cavernous sinus, as shown by preoperative MRI, or the medial wall of the cavernous sinus, as shown in the intraoperative findings. The extent of the resection was classified as total when tumor was completely removed, subtotal when more than 95% of tumor was removed, and partial when less than 95% of tumor was removed, which was based on both intraoperative findings and immediate postoperative MRI results taken within 2 days after surgery.

Biochemical Measurements

In all patients, IGF-1 levels and nadir GH levels during the OGTT were measured before surgery; 1 week, 6 months, and 12 months after surgery; and thereafter annually (23). An OGTT was used to determine the remission status at each time point. The association of IGF-1 level with remission status as determined by the OGTT was evaluated. For the second analysis, 270 patients who achieved GH normalization were divided into 2 groups: an OGTT/IGF-1 concordant group with normalized IGF-1 levels and an OGTT/IGF-1 discordant group with sustained elevation of IGF-1 levels. Then, the factors associated with OGTT/IGF-1 discordance were analyzed.

An immunoradiometric assay (Cis Bio International, Gif-sur-Yvette, France) was used to measure GH levels. This assay had a sensitivity of 0.03 mIU/mL; the interassay coefficient of variation (CV) was 3.8% to 5.0%, and the within-assay CV was 1.3% to 2.1%. The World Health Organization international standard 98/574 was used to classify GH levels. The IGF-1 Next immunoradiometric assay computed tomography system (Biocode Hycel, Liège, Belgium) was used to measure IGF-1 levels, calibrated to the World Health Organization international standard 02/254. The minimum detectable IGF-1 concentration was 1.25 ng/L, the interassay CV was 7.4% to 9.1%, and the within-assay CV was 2.6% to 4.4%.

When GH levels were below 1.0 ng/mL during the OGTT after surgery, it was defined as an endocrine remission. The uIGF-1 level was defined as the percentile of the upper limit of the normal IGF-1 range.

Statistical Analysis

For statistical analyses, IBM SPSS Statistics software (version 25.0; IBM, Armonk, NY, USA) was used. The OGTT was performed 1 week, 6 months, 12 months, 24 months, and 36 months after surgery to determine the status of endocrine remission at each time point. Receiver operator characteristic (ROC) analysis was performed to investigate whether the endocrine remission status of 1 week and 6, 12, 24, and 36 months after surgery can be judged by uIGF-1 measured at each time point. The chi-squared test, Fisher’s exact test, and Student’s t-test were performed to demonstrate statistical significance of any differences. Univariate and multivariate logistic regression analyses were performed to identify possible factors associated with OGTT/IGF-1 discordance. Variates with a P-value less than 0.1 were included in the multivariable model. Whether the model of regression analysis is a good fit to the data was verified using the Hosmer and Lemeshow test.

Results

Clinical and Biochemical Characteristics

We identified 320 adult patients with acromegaly (169 females and 151 males) who underwent TSS of their GH-secreting adenomas. The mean age of the patients was 44.4 ± 11.4 years, and the mean follow-up duration was 30.3 ± 12.6 months. The mean body mass index (BMI) was 25.1 ± 3.0 kg/m². At the time of diagnosis, 81 patients (25.3%) had diabetes mellitus, and 103 patients (32.2%) had hypertension. Cavernous sinus invasion was identified before and during surgery in 117 patients (36.6%). The mean maximal tumor diameter was 16.8 ± 7.8 mm. Total resection was achieved in 281 patients (87.8%). In 270 patients (84.3%), GH levels were below 1.0 ng/mL during the OGTT after surgery, which signified endocrine remission. The mean preoperative nadir GH and IGF-1 levels were 17.8 ± 36.7 ng/mL and 732.3 ± 208.4 ng/mL, respectively (237.9 ± 68.4% of the upper normal limit). Details of the clinical and biochemical characteristics of the patients are summarized in Table 1. In a remission group with 270 patients, 243 patients (90.0%) achieved remission on 1-week postoperative OGTT and 265 patients (98.1%) achieved remission by 6 months. Only 2 patients (0.7%) have ever showed insufficient suppression of GH (nadir GH >1.0 ng/mL) on OGTT during follow-up until they finally achieved remission on the last OGTT. In a nonremission group, failure to achieve remission by surgery alone was determined in a more delayed form. In particular, 9 patients (18.0%) showed inconstant OGTT results during follow-up.

Most patients who experienced the subtotal removal of their tumors underwent adjuvant Gamma Knife radiosurgery (GKS). A somatostatin analogue was discontinued for 3 months prior to GKS or not started until after GKS (24, 25).

Remission Status and IGF-1 Level at Each Time Point

One week after surgery, endocrine remission was achieved in 80.6% of patients whose GH levels were below 1.0 ng/mL during the OGTT, while only 31.7% showed simultaneous normalization of IGF-1 levels (Fig. 1). The proportion of patients experiencing remission with consistent OGTT results and IGF-1 levels continued to increase over time, reaching 86.6% by 36 months postoperatively. The proportion of patients with discordant OGTT results and IGF-1 levels was 53.3% by 1 week postoperatively, which decreased to 16.2% by 6 months after surgery.

Reliability of Postoperative IGF-1 Measurement

To determine endocrine remission after surgery, OGTT was performed at 1 week and 6, 12, 24, and 36 months after surgery. And ROC analysis was performed to investigate whether uIGF-1 levels measured at the same time points would show consistent results with OGTT (Table 2). Except for 1 week after surgery, the area under the curve (AUC) of uIGF-1 was above 0.8 at all time periods (Table 2). Specifically, uIGF-1 by 6 months postoperatively had an AUC of greater than or equal to 0.9, which implies outstanding discrimination (Fig. 2). The cutoff value of uIGF-1 for predicting endocrine remission was determined using the maximum Youden’s J (sensitivity + specificity – 1.0). By 6 months after surgery, the cutoff value of uIGF-1 was determined to be 114.922%, its sensitivity was 0.938, and its specificity was 0.811.

OGTT/IGF-1 Discordance

Of the 320 patients analyzed, 270 patients finally achieved a nadir GH level of less than 1.0 ng/mL during the OGTT after surgical treatment alone. Of these patients, we observed IGF normalization in 108 (40.0%), 224 (83.0%), 242 (89.6%), 247 (91.5%), and 250 (92.6%) patients at 1 week, and at 6, 12, 24, and 36 months after surgery, respectively. However, in 20 patients (7.4%), IGF-1 levels did not normalize until 36 months postoperatively (Fig. 3). The mean duration from surgery to IGF-1 normalization was 4.7 months.

We grouped the 270 patients who achieved endocrine remission into 250 OGTT/IGF-1 concordant patients and 20 OGTT/IGF-1 discordant patients. The clinical and biochemical characteristics of these 2 groups were compared (Table 3). Sex, age, BMI, presence of diabetes mellitus, presence of hypertension, maximal tumor diameter, and preoperative hormone levels did not differ between the 2 groups. Preoperative uIGF-1 levels were higher in OGTT/IGF-1 discordant patients than in concordant patients, although this result was not statistically significant (P = .097). However, OGTT/IGF-1 discordant patients exhibited more frequent invasion of the cavernous sinus (P = .012) and subtotal resection (P = .021) than concordant patients.

Multivariate and univariate logistic regression analyses were performed to identify possible factors associated with OGTT/IGF-1 discordance. Multivariate logistic regression analyses were performed on the variables with P < .10 in the univariate analysis. Cavernous sinus invasion, extent of resection, and preoperative uIGF-1 were included in the multivariate logistic regression analysis while only cavernous sinus invasion and extent of resection showed statistical significance (Table 4). In patients who achieved remission, the possibility of failed IGF-1 normalization was increased by 3.2-fold (95% CI 1.218-8.274, P = .018) by the presence of cavernous sinus invasion. In addition, the possibility of failed IGF-1 normalization was increased by 11.8-fold (95% CI 1.429-96.994, P = .022) when the tumors were not completely removed (Table 4).

Then, we performed subgroup analysis focusing on OGTT/IGF-1 discordant patients, in which clinical and biochemical characteristics between 20 patients with normal OGTT/high IGF-1 and 16 patients with high OGTT/normal IGF-1 were compared (Table 5). Patients with high OGTT/normal IGF-1 more likely associated with unfavorable prognostic characteristics such as bigger tumor size, incomplete tumor resection as well as lower preoperative uIGF-1 levels and lower body mass index. Also, tumors with higher Hardy grade and presence of cavernous sinus invasion were more prevalent in patients with high OGTT/normal IGF-1 although it failed to obtain statistical significance.

Patient Group With Remnant Tumors

A total of 39 (12.2%) patients failed to achieve the complete removal of their tumors by TSS: subtotal resection in 38 patients and partial resection in 1 patient. In 28 patients (71.8%), remnant tumors were observed by immediate postoperative MRI. In the other 11 patients (28.2%), remnant tumors were not identified by immediate postoperative MRI but later became visible during follow-up.

Of the 39 patients with remnant tumors, 34 patients (87.2%) eventually underwent adjuvant GKS, followed by the administration of a somatostatin analogue. GKS was performed before 6 months in 3 patients, between 6 and 12 months in 22 patients, and after 12 months in 9 patients. Additionally, 5 patients (12.8%) did not undergo adjuvant treatment.

By 6 months after surgery, OGTT and IGF-1 levels were compared for their power in predicting the presence of remnant tumors in 36 patients after excluding 3 patients who underwent GKS within 6 months. In 24 patients (66.7%), both OGTT and IGF-1 levels were indicative of incomplete tumor removal with sustained acromegalic condition. Overall, sustained acromegaly was confirmed by either OGTT or IGF-1 measurement in 31 patients (86.1%). However, in 5 patients (13.9%), both OGTT and IGF-1 levels confirmed that the acromegaly had been resolved. The positive predictive value (PPV) for remnant tumors using nadir GH levels during the OGTT was 77.8%, while the PPV of IGF-1 levels was 75.0% (Table 6).

Discussion

GH is secreted directly from the pituitary, whereas IGF-1 is secreted from the liver or other GH-stimulated tissue. Therefore, in patients with acromegaly who underwent surgical treatment, the monitoring of GH levels is a more direct way to evaluate the disease state than the monitoring of IGF-1 levels (19). However, because of the pulsatile nature of GH secretion, random measurements of serum GH levels do not properly reflect the GH-secreting activity of the tumor. Therefore, patients with acromegaly who have undergone surgical treatment should be monitored for tumor endocrine activity using nadir GH measurements during the OGTT, not using random GH measurements. The definition of endocrine remission of acromegaly has changed over time. Nadir GH levels less than 1.0 ng/mL during the OGTT were believed to reflect a postoperative cure in 2000 (26). In later papers, remission criteria for acromegaly were more stringently defined as nadir GH levels less than 0.4 ng/mL during the OGTT (7, 27). However, our institution has adopted a nadir GH level less than 1.0 ng/mL as a remission criterion for patients with acromegaly based on previous reports (23, 28-31).

On the other hand, performing an OGTT is more cumbersome than measuring IGF-1 levels because overnight fasting and multiple blood sampling are required. In addition, the reliability of nadir GH levels is being questioned, especially in patients administered a somatostatin analogue (8, 27, 32, 33). For these reasons, many institutions prefer measuring IGF-1 levels than performing an OGTT to evaluate hormonal activity in patients with acromegaly. Nevertheless, it is obvious that the OGTT is considered the gold standard for the diagnosis of and determination of remission in patients with acromegaly. In this context, we investigated whether the OGTT could be replaced by IGF-1 measurements for the determination of disease remission and detection of recurrence after surgical treatment in patients with acromegaly.

ROC analyses based on a comparison of OGTT and IGF-1 levels revealed that the AUC exceeded 0.8 at all time points except 1 week postoperatively, suggesting that IGF-1 is an excellent biomarker of endocrine remission. In particular, the highest AUC (0.928) was obtained 6 months after surgery. IGF-1 levels were not reliable 1 week postoperatively, as the AUC of IGF-1 at this time was merely 0.691. Therefore, our results confirm that IGF-1 should not be used as an early postoperative predictor for endocrine remission. A longer time is reportedly required for IGF-1 stabilization after surgery (12, 34), although the half-life of IGF-1 is only several hours.

We also performed a subgroup analysis of 270 patients with acromegaly who achieved remission after surgery alone, which was verified by the OGTT results. Less than one-half of these patients (108 patients, 40.0%) exhibited normalization of IGF levels within 1 week after surgery. But the number of patients with normalized IGF levels abruptly increased and reached 83.0% (224 patients) by 6 months after surgery. Overall, 250 patients (92.6%) showed consistent normalization of IGF-1 levels within 36 months postoperatively. Thus, we conclude that IGF-1 can be used as a reliable predictor of endocrine remission 6 months after surgery. In our study, the average time required for IGF-1 normalization was 4.7 months, which is shorter than that reported in previous studies. Shin et al. (34). reported that the mean duration for IGF-1 normalization after surgical treatment was 10 months in a retrospective, observational study of 46 patients with acromegaly. In their study, the mean follow-up duration was 61 months. Serum IGF-1 levels were measured before surgery, within 3 days after surgery, at 3-month intervals for 1 year, and once or twice yearly thereafter. Espinosa-de-los-Monteros et al. (12). reported that in 53 patients with acromegaly who underwent surgical treatment, it took 11.6 ± 5.7 months for IGF-1 normalization. Bacigaluppi et al. (35). showed that early postoperative IGF-1 levels (1 month) are significantly lower in patients achieving GH and IGF-1 normalization. And early postoperative IGF-1 normalization occurred in 53.8% of patients who were cured by surgery alone. These findings are in line with the observation of our study in which normalization of IGF-1 is achieved much earlier after surgical treatment of acromegaly than previously known.

We also assessed the 20 patients (7.4%) whose IGF-1 levels never normalized throughout the follow-up period. There have been several studies on the discordance between OGTT results and IGF-1 levels in the literature. When a nadir GH level less than 1.0 ng/mL during the OGTT was defined as the cutoff, Machado et al (30). reported the prevalence of discordance to be 13.7% at diagnosis (7 out of 51 patients), 13.8% after the surgical procedure (8 out of 58 patients), and 33.3% during treatment with octreotide (14 out of 42 patients). Their patients were evaluated using 2 consecutive hormonal assessments at the time of diagnosis, 3 months after surgical treatment, and after at least 6 months of treatment with octreotide. These authors did not find a significant difference in sex, age, or treatment modality between the discordant and concordant groups. Brzana et al. (36). retrospectively reviewed patients with acromegaly who showed elevated IGF-1 levels but normal GH levels after surgery. In their study, 13 of 54 (24%) patients were found to have OGTT/IGF-1 discordance during their final evaluation. These authors also failed to show significant differences in age, sex, BMI, presence of diabetes mellitus, and adrenal insufficiency between the discordant and concordant groups. In another retrospective study by Zhang et al. (37), OGTT/IGF-1 discordance was noted in 31 (60.8%) of 51 patients who had achieved endocrine remission after surgery alone. Their discordant group had significantly higher mean BMI and IGF-1 levels than the concordant group. The prevalence of OGTT/IGF-1 discordance reported in this paper was much higher than that in previous reports because hormone levels were measured only once and too soon after surgery. These authors assessed discordance by hormone levels measured 3 months after surgery, but 3 months is insufficient for IGF-1 levels to have normalized after surgical treatment (12, 33).

In this study, we divided 270 patients who achieved remission after surgery alone into 250 OGTT/IGF-1 concordant patients and 20 discordant patients, and a regression analysis between the 2 groups was performed. We found that the risk of failed IGF-1 normalization increased 3.2- and 11.8-fold by the invasion of the cavernous sinus and the presence of a remnant tumor after surgery, respectively. Cavernous sinus invasion is the most significant prognostic factor that hinders radical surgical resection during TSS. Thus, IGF-1/OGTT discordance can be indicative of incomplete surgical resection. Our observation suggests that patients who exhibit failure of IGF-1 normalization despite sufficient GH suppression during the OGTT should be carefully assessed for possible remnant or recurrent tumors. In the subgroup comparative analysis focusing on OGTT/IGF-1 discordant patients, in which clinical and biochemical characteristics between 20 patients with normal OGTT/high IGF-1 and 16 patients with high OGTT/normal IGF-1 were compared, patients with high OGTT/normal IGF-1 were more likely associated with unfavorable prognostic characteristics such as bigger tumor size, incomplete tumor resection. This may implicate that OGTT may reflect incomplete treatment caused by many of unfavorable characteristics more sensitively.

In the subgroup analysis of 39 patients with obvious remnant tumors, we evaluated which parameter (OGTT results or uIGF-1 levels) is superior for predicting a remnant GH-secreting adenoma with sustained acromegaly. Because our results indicated that 4.7 months is required for IGF-1 stabilization and that most patients did not undergo GKS until 12 months after resection, IGF-1 measurement and OGTT were evaluated 6 months after surgery and compared in terms of their predictive power. Five patients with obvious remnant tumors presented with endocrine remission, as indicated by both normalization of IGF-1 levels and sufficient GH suppression during the OGTT. There was no significant difference between the PPV of the OGTT and that of the IGF-1 measurements. Surprisingly, 13.9% of patients with obvious remnant tumors still showed endocrine outcomes indicative of remission suggested by both the OGTT and IGF-1 measurement. Further, 19.4% of patients showed discordance between OGTT and IGF-1 findings, and thus sustained acromegaly was demonstrated by either OGTT or IGF-1 levels in up to 86.1% of patients with incomplete tumor removal. We believe that both OGTT and IGF-1 levels are important to monitor a tumor’s biochemical activity in patients with remnant tumors after surgery.

Our study is limited by its retrospective study design. Even though it is the largest study of GH and IGF-1 discordance to date, it utilized a small sample size. Finally, not all factors affecting serum IGF-1 levels were considered. Nutrient deprivation, malnutrition, anorexia nervosa, liver disease, hypothyroidism, poorly controlled insulin-dependent diabetes mellitus, and use of oral estrogen can lower serum IGF-1 levels. Additionally, adolescence, pregnancy, and hyperthyroidism can falsely elevate serum IGF-1 levels (38). Nevertheless, our study strongly supports that IGF-1 level is a reliable tool to monitor acromegaly status after surgery. Less than 114.9% of uIGF-1 at postoperative 6 months reliably predicts endocrine remission.

Abbreviations

Abbreviations
 
• AUC

  area under the curve

 
• BMI

  body mass index

 
• CV

  coefficient of variation

 
• GKS

  Gamma Knife radiosurgery

 
• GH

  growth hormone

 
• IGF-1

  insulin-like growth factor-1

 
• MRI

  magnetic resonance imaging

 
• OGTT

  oral glucose tolerance test

 
• PPV

  positive predictive value

 
• ROC

  receiver operator characteristic

 
• TSS

  transsphenoidal surgery

Acknowledgments

The authors thank Juyoon Park, RN, MPH, OCN, Min Kyeong Jang, PhD, RN, KOAPN, Sung Ja Kang, RN, Sujin Ryu, RN, and Yong Jun Jang, RN for their tremendous effort in performing the endocrinological tests and data acquisition for such a long follow-up duration.

Financial Support: This study was supported by a faculty research grant of Yonsei University College of Medicine (6-2020-0224) and the Team Science Award Grant of Yonsei University College of Medicine (6-2021-0009).

Additional Information

Disclosures: The authors have nothing to disclose.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

References