Browse the corpus

Association Between p-STAT3/STAT3 Expression and Knosp Grading in Growth Hormone Pituitary Tumors. BACKGROUND AND OBJECTIVES: Growth hormone-secreting pituitary adenomas (GH-PAs) that invade the structures surrounding the pituitary gland are a big problem for clinical professionals working in the field. People have tried to find a reliable biomarker for this. The STAT3 and phosphorylated STAT3 (p-STAT3) proteins are increased in many cancers and help cells move and invade. This study looked at the link between p-STAT3/STAT3 levels and the Knosp grading system for GH-PAs. METHODS: A total of 71 patient GH-PAs were obtained between 22 January 2016 and 23 January 2022. An immunohistochemical analysis was conducted to assess the expression of p-STAT3/STAT3 and keratin CAM5.2 in tumor samples. p-STAT3/STAT3 was also used to evaluate the predictive value of tumor invasion propensity and patient prognosis. RESULTS: Tumor invasion, maximum diameter, macroadenoma, high T2-weighted intensity signal, cystic degeneration, low remission, and resection rates were found to be associated with high p-STAT3 expression. A large adenoma, a high T2-weighted intensity signal, cystic change, high insulin-like growth factor 1 levels, sparsely granular adenomas, high p-STAT3 expression, and a low resection rate of postoperative tumor were all associated with increased tumor aggressiveness. However, there was no link between Ki-67 and STAT3 expression and tumor invasion. Using MRI to diagnose the invasiveness of GH-PAs, it was found that p-STAT3 was more effective than sparsely granulated. CONCLUSION: The findings of this study indicate that the elevated expression of p-STAT3 may serve as a potential biomarker for the prediction of the invasiveness and clinical outcome of GH-PAs.

All patients gave their consent, and the study was approved. This study was performed at the Department of Neurosurgery between 23 January 2016 and 22 January 2022. The study was performed at one hospital. We got a lot of medical records, images, and samples. The study included patients who met the following criteria: (1) had acromegaly, (2) underwent initial surgery, and (3) had presurgical MRI. People were excluded if they (1) had previous radiotherapy or surgery to the pituitary, (2) had previous treatment for acromegaly (Pegvisomant, somatostatin receptor ligands, GH receptor antagonist, or cabergoline) before surgery; and (3) were unable to access clinical, laboratory, or imaging data. A total of 87 patients who met the inclusion criteria and did not present any exclusion criteria were included in this study. However, 16 cases were excluded because of the absence of essential clinical data and/or laboratory and/or imaging data. This resulted in 71 GH-PAs being available for analysis (Supplemental Digital Content 1 [http://links.lww.com/NEU/E757]). Subsequently, all patients underwent an assessment to determine the level of tumor aggressiveness, and all of them underwent Keratin CAM5.2 and p-STAT3/STAT3 immunohistochemical staining. In this retrospective study, we examined the correlation between various factors, including age, sex, body mass index (BMI), the presence of cystic changes, the Ki-67 labeling index, preoperative and postoperative MRI with and without gadolinium injection, and the correlation between these factors and the GH-PAs.

ning. In this retrospective study, we examined the correlation between various factors, including age, sex, body mass index (BMI), the presence of cystic changes, the Ki-67 labeling index, preoperative and postoperative MRI with and without gadolinium injection, and the correlation between these factors and the GH-PAs. In patients who had undergone neurosurgery, hormonal remission was defined by postoperative normalization of age-adjusted insulin-like growth factor 1 (IGF-1) levels and random GH levels of <1 ng/mL or a nadir GH of <0.4 ng/mL during oral glucose tolerance test at least 3 months after surgery.12-15 As previously stated,16 the IGF-1 index (calculated as the IGF-1 value divided by the maximal normal value for the assay) was also determined to account for potential assay and range variability. The procedure for the selection of GH-PA tissue is outlined in Supplemental Digital Content 2 (http://links.lww.com/NEU/E758). The Partners Human Research Committee of the aforementioned hospital granted approval for the acquisition of consent, and the study was conducted in accordance with the Declaration of Helsinki. All patients provided informed consent for the use of their data and surgical specimens after the histopathological diagnosis.

The Partners Human Research Committee of the aforementioned hospital granted approval for the acquisition of consent, and the study was conducted in accordance with the Declaration of Helsinki. All patients provided informed consent for the use of their data and surgical specimens after the histopathological diagnosis. Before undergoing surgery, all patients underwent a preoperative MRI scan. A retrospective evaluation of the T2-weighted intensity (T2WI) signal on the preoperative images was conducted. An adenoma was classified as invasive if it extended into the sphenoid sinuses and/or invaded the cavernous sinus, according to grade 3 or 4 of the Knosp classification. Furthermore, it was regarded as noninvasive and had a Knosp grading of 0-2 on T1-weighted or postcontrast T1-weighted MRIs.17,18 The percentage of tumor resection, as previously described,19 was based on the MRI results of the postoperative follow-up. The STAT3 and p-STAT3 (Tyr705) antibodies were sourced from Cell Signaling Technology and used at dilutions of 1:500 and 1:200, respectively. The Keratin CAM5.2 antibody and Pit-1 antibody were obtained from Zhongshan Jinqiao Biotechnology Co., LTD. Tissue samples embedded in paraffin were sectioned at a thickness of 4 μm and subsequently stained with hematoxylin and eosin for routine histopathological examination and diagnosis.

The STAT3 and p-STAT3 (Tyr705) antibodies were sourced from Cell Signaling Technology and used at dilutions of 1:500 and 1:200, respectively. The Keratin CAM5.2 antibody and Pit-1 antibody were obtained from Zhongshan Jinqiao Biotechnology Co., LTD. Tissue samples embedded in paraffin were sectioned at a thickness of 4 μm and subsequently stained with hematoxylin and eosin for routine histopathological examination and diagnosis. The staining intensity values of 0 (no staining), 1 (mild staining), 2 (moderate staining), and 3 (intense staining) were used to assess the level of expression, respectively. Scores of 0, 1, 2, and 3 were assigned based on the percentages of stained cells falling into the categories of <10, 10 to 25, 25 to 50, and more than 50%, respectively. A score of >2 indicated positive expression, whereas a score of 2 or less indicated negative expression.20 Some of or all the data sets generated and/or analyzed during the current study are not publicly available, but are available from the corresponding author on reasonable request. Statistical analysis was conducted using IBM SPSS Statistics for Windows, version 23 (IBM Corp.). The numerical data are presented as the mean.

The staining intensity values of 0 (no staining), 1 (mild staining), 2 (moderate staining), and 3 (intense staining) were used to assess the level of expression, respectively. Scores of 0, 1, 2, and 3 were assigned based on the percentages of stained cells falling into the categories of <10, 10 to 25, 25 to 50, and more than 50%, respectively. A score of >2 indicated positive expression, whereas a score of 2 or less indicated negative expression.20 Some of or all the data sets generated and/or analyzed during the current study are not publicly available, but are available from the corresponding author on reasonable request. Statistical analysis was conducted using IBM SPSS Statistics for Windows, version 23 (IBM Corp.). The numerical data are presented as the mean. Parametric data are described using the mean and SD, whereas nonparametric data are described using the median and the 25th and 75th percentiles. Categorical data are presented as absolute numbers (n) and percentages (%). In instances where appropriate, the nonparametric Mann–Whitney U test or 1-way analysis of variance was used to statistically compare quantitative data. In statistical comparisons of dichotomous data, the Fischer’s exact test or the χ2 test was used. McNemar identified discrepancies between p-STAT3, sparsely granulated (SG), and MRI results, whereas the Kappa coefficient quantified the consistency of diagnostic findings. The strength of the Kappa coefficient was classified as follows: weak (κ < 0.4), general (κ between 0.4 and 0.7), and strong (κ > 0.7). In addition, we used receiver operating characteristic (ROC) curves for diagnostic testing. All statistical tests were two-sided and considered significant at P-values <.05.

Before undergoing surgery, all patients underwent a preoperative MRI scan. A retrospective evaluation of the T2-weighted intensity (T2WI) signal on the preoperative images was conducted. An adenoma was classified as invasive if it extended into the sphenoid sinuses and/or invaded the cavernous sinus, according to grade 3 or 4 of the Knosp classification. Furthermore, it was regarded as noninvasive and had a Knosp grading of 0-2 on T1-weighted or postcontrast T1-weighted MRIs.17,18 The percentage of tumor resection, as previously described,19 was based on the MRI results of the postoperative follow-up.

The STAT3 and p-STAT3 (Tyr705) antibodies were sourced from Cell Signaling Technology and used at dilutions of 1:500 and 1:200, respectively. The Keratin CAM5.2 antibody and Pit-1 antibody were obtained from Zhongshan Jinqiao Biotechnology Co., LTD. Tissue samples embedded in paraffin were sectioned at a thickness of 4 μm and subsequently stained with hematoxylin and eosin for routine histopathological examination and diagnosis. The staining intensity values of 0 (no staining), 1 (mild staining), 2 (moderate staining), and 3 (intense staining) were used to assess the level of expression, respectively. Scores of 0, 1, 2, and 3 were assigned based on the percentages of stained cells falling into the categories of <10, 10 to 25, 25 to 50, and more than 50%, respectively. A score of >2 indicated positive expression, whereas a score of 2 or less indicated negative expression.20

Statistical analysis was conducted using IBM SPSS Statistics for Windows, version 23 (IBM Corp.). The numerical data are presented as the mean. Parametric data are described using the mean and SD, whereas nonparametric data are described using the median and the 25th and 75th percentiles. Categorical data are presented as absolute numbers (n) and percentages (%). In instances where appropriate, the nonparametric Mann–Whitney U test or 1-way analysis of variance was used to statistically compare quantitative data. In statistical comparisons of dichotomous data, the Fischer’s exact test or the χ2 test was used. McNemar identified discrepancies between p-STAT3, sparsely granulated (SG), and MRI results, whereas the Kappa coefficient quantified the consistency of diagnostic findings. The strength of the Kappa coefficient was classified as follows: weak (κ < 0.4), general (κ between 0.4 and 0.7), and strong (κ > 0.7). In addition, we used receiver operating characteristic (ROC) curves for diagnostic testing. All statistical tests were two-sided and considered significant at P-values <.05.

Table 1 presents the demographic and demographic characteristic data for a total of 71 patients exhibiting clinical symptoms of acromegaly. The study group comprised 29 males and 42 females, with a mean age of 42.46 years (±1.57 years) and a mean BMI of 24.96 (±2.96). The mean levels of IGF-1 and xULN-IGF-1 were found to be 418.26 ng/mL (±188.96 ng/mL) and 1.20 (±0.54), respectively. The GH levels were categorized into 3 groups: >50 ng/mL, 20 to 50 ng/mL, and ≤20 ng/mL. These were observed in 32 (45.07%), 21 (29.58%), and 18 (25.35%) tumor samples, respectively. The MRI findings (Figure 1) revealed that the signals of T2WI hyper-, iso-, and hypointense were present in 24 (33.80%), 16 (22.54%), and 31 (43.66%) tumor samples, respectively. Of the total number of patients, 67 (94.37%) exhibited macroadenomas, with a median maximal tumor diameter of 2.65 cm (±1.47 cm). In addition, 41 patients displayed cystic changes as observed through MRI (Figure 221). Invasion (Knosp grade 3 or 4) was observed in 50 patients (70.42%) (Table 1 and Figure 3). It is noteworthy that 16 patients (22.54%) achieved remission after surgery. In addition, the extent of postoperative resection of the tumor was evaluated, identifying gross total resection, subtotal resection, and partial resection in 23 (32.29%), 16 (22.54%), and 32 (45.07%) tumor samples, respectively (Table 1 and Supplemental Digital Content 2 [http://links.lww.com/NEU/E758]). In Ki-67 evaluation, 9 adenomas (12.67%) exhibited positivity (>3%), whereas 69 (87.32%) demonstrated negativity (≤3%). Individual data are available in Supplemental Digital Content 1 (http://links.lww.com/NEU/E757).

mor samples, respectively (Table 1 and Supplemental Digital Content 2 [http://links.lww.com/NEU/E758]). In Ki-67 evaluation, 9 adenomas (12.67%) exhibited positivity (>3%), whereas 69 (87.32%) demonstrated negativity (≤3%). Individual data are available in Supplemental Digital Content 1 (http://links.lww.com/NEU/E757). Baseline Clinical Data and Tumor Characteristics of 71 Patients With Growth Hormone–Secreting Pituitary Adenomas BMI, Body mass index; DG-A, densely granulated adenoma; F, female; GH, growth hormone; GTR, gross total resection; Hyper, hyperintense; I-A, intermediate or transitional adenoma; Iso, isointense; Hypo, hypointense; IGF-1, insulin-like growth factor 1; Macro, macroadenoma; M, male; Micro, microadenoma; MRI, magnetic resonance imaging; NC-A, Keratin-negative adenoma; PR, partial resection; SG-A, sparsely granulated adenoma; STR, subtotal resection; T2WI, T2-weighted intensity; ULN, upper limit normal range. There are n (%) categorical data displayed. The objective of this study is to assess the intensity of growth hormone–secreting pituitary adenomas on T2-weighted MRIs using a visual assessment method. A, On T2 MRI, the tumor presents as a lighter shade (hyperintense T2 intensity) than gray matter; B, On T2 MRI, the tumor displays an intensity range (isointense T2 intensity) between white and gray matter; C, On T2 MRI, the tumor presents as a darker shade (hypointense T2 intensity) than gray matter.

ssessment method. A, On T2 MRI, the tumor presents as a lighter shade (hyperintense T2 intensity) than gray matter; B, On T2 MRI, the tumor displays an intensity range (isointense T2 intensity) between white and gray matter; C, On T2 MRI, the tumor presents as a darker shade (hypointense T2 intensity) than gray matter. Invasive and noninvasive GH-PAs were observed to display varying degrees of staining intensity for p-STAT3/STAT3 and CAM5.2 markers.21 A, The invasive GH-PA had invaded the cavernous sinus and completely enclosed the internal carotid artery (Knosp grade 4). B, A noninvasive GH-PA, confined to the intrasellar region, was observed (Knosp grade 2). C, The p-STAT3 staining was positive for the invasive GH-PA. D, The p-STAT3 staining was negative for the noninvasive GH-PA. E, The STAT3 staining was positive for the GH-PA. F, The STAT3 staining was positive for the noninvasive GH-PA. G, The staining of the SG-A demonstrated a dot pattern in ≥70% of the cells. H, Densely granulated adenoma staining demonstrated a perinuclear or transitional pattern with a dot pattern of ≤8%. I, Invasive GH-PA exhibited cytoplasm that was chromophobic to pale eosinophilic (H&E). J, Noninvasive GH-PA displayed cytoplasm that was chromophobic to dense eosinophilic (H&E). GH-PA, growth hormone–secreting pituitary adenoma; H&E, hematoxylin and eosin; p-STAT3, phosphorylated STAT3.

ern with a dot pattern of ≤8%. I, Invasive GH-PA exhibited cytoplasm that was chromophobic to pale eosinophilic (H&E). J, Noninvasive GH-PA displayed cytoplasm that was chromophobic to dense eosinophilic (H&E). GH-PA, growth hormone–secreting pituitary adenoma; H&E, hematoxylin and eosin; p-STAT3, phosphorylated STAT3. Types of cystic lesions in growth hormone–secreting pituitary adenomas. A, Presence of scattered, spotty cystic lesions; B, Some scattered, cord cystic lesions; C, Some plaque-like cystic lesions; D, Multiple cystic lesions have merged into large cystic lesions. Immunohistochemical staining was conducted using p-STAT3/STAT3 and Keratin CAM5.2 in 71 GH-PAs. The results demonstrated that p-STAT3 was positive in 50 cases and negative in 21 cases (Figure 2 and Table 2), whereas STAT3 was positive in 69 cases and negative in only 2 cases (Figure 2 and Supplemental Digital Content 1 [http://links.lww.com/NEU/E757]). Furthermore, the CAM5.2 expression pattern identified 4 pathogenic subtypes: densely granular adenomas (n = 35; 49.29%), SG adenomas (SG-As) (n = 21; 29.58%), intermediate or transitional adenoma (n = 7; 9.82%), and Keratin-negative adenoma (n = 11.27%) (Table 2, Figure 2 and Supplemental Digital Content 2 [http://links.lww.com/NEU/E758]).22 Immunoreactivity of p-STAT3/STAT3 and CAM5.2 in Growth Hormone–Secreting Pituitary Adenomas DG-A, densely granulated adenoma; I-A, intermediate or transitional adenoma; NC-A, keratin-negative adenoma; p-STAT3, phosphorylated STAT3; SG-A, sparsely granulated adenoma.

Immunohistochemical staining was conducted using p-STAT3/STAT3 and Keratin CAM5.2 in 71 GH-PAs. The results demonstrated that p-STAT3 was positive in 50 cases and negative in 21 cases (Figure 2 and Table 2), whereas STAT3 was positive in 69 cases and negative in only 2 cases (Figure 2 and Supplemental Digital Content 1 [http://links.lww.com/NEU/E757]). Furthermore, the CAM5.2 expression pattern identified 4 pathogenic subtypes: densely granular adenomas (n = 35; 49.29%), SG adenomas (SG-As) (n = 21; 29.58%), intermediate or transitional adenoma (n = 7; 9.82%), and Keratin-negative adenoma (n = 11.27%) (Table 2, Figure 2 and Supplemental Digital Content 2 [http://links.lww.com/NEU/E758]).22 Immunoreactivity of p-STAT3/STAT3 and CAM5.2 in Growth Hormone–Secreting Pituitary Adenomas DG-A, densely granulated adenoma; I-A, intermediate or transitional adenoma; NC-A, keratin-negative adenoma; p-STAT3, phosphorylated STAT3; SG-A, sparsely granulated adenoma. For this purpose, the entire cohort of 71 patients was subjected to evaluation. Table 3 presents the associations between tumor invasiveness and patient characteristics, laboratory data, and imaging findings. The results demonstrated no statistically significant differences between the invasive GH-PA group and the noninvasive GH-PA group with regard to sex, age distribution, BMI, Ki-67, GH basal level, and GH level after surgery (Table 3). The maximum diameter of the tumor before surgery was significantly larger in the invasive GH-PA group (3.09 ± 1.51 cm) compared with the noninvasive GH-PA group (1.68 ± 0.77 cm; P = 0). In addition, the invasive GH-PA group exhibited a higher prevalence of macroadenoma (100% vs 80.95%; P = .006) and higher levels of IGF-1 (4 basal levels of 0.0125 ng/mL (313.50-548.80) vs 304.90 ng/mL (263.85-429.40) were associated with tumor invasiveness (P = .009). Furthermore, the group of aggressive GH-PAs demonstrated a notable increase in the prevalence of cystic variability (70.00% vs 28.57%; P = .001) and a considerable number of hyperintense signals in T2WI (44.00% vs 9.52%; P = .005) in comparison with the nonaggressive GH-PA group. As anticipated, patients with noninvasive tumors exhibited low IGF-1 levels (89.30 ng/mL [121.35-277.15] vs 352.10 ng/mL [234.40-453.40]; P = .009). In the postoperative period, a higher rate of surgical remission (38.10% vs 16.00%; P = .042, Table 3) and total resection percentage (80.95% vs 12.00%; P = 0, Table 3) were observed in patients with noninvasive tumors compared with those with invasive tumors.

[121.35-277.15] vs 352.10 ng/mL [234.40-453.40]; P = .009). In the postoperative period, a higher rate of surgical remission (38.10% vs 16.00%; P = .042, Table 3) and total resection percentage (80.95% vs 12.00%; P = 0, Table 3) were observed in patients with noninvasive tumors compared with those with invasive tumors. Comparison of Patient Characteristics Based on Tumor Invasiveness in the Clinical, Laboratory, and Imaging Domains BMI, Body mass index; DGA, densely granulated adenomas; F, female; GH, growth hormone; GTR, gross total resection; Hyper, hyperintense; Iso, isointense; Hypo, hypointense; IGF-1, insulin-like growth factor 1; M, male; Macro, macroadenoma; Micro, microadenoma; MRI, magnetic resonance imaging; PR, partial resection; p-STAT3, phosphorylated STAT3; SGA, sparsely granulated adenomas; STR, subtotal resection; T2WI, T2-weighted intensity; ULN, upper limit normal range. There are n (%) categorical data displayed. Fisher's exact test for comparison between the invasive group vs the noninvasive Group. Bold values indicate statistically significant P values.

BMI, Body mass index; DGA, densely granulated adenomas; F, female; GH, growth hormone; GTR, gross total resection; Hyper, hyperintense; Iso, isointense; Hypo, hypointense; IGF-1, insulin-like growth factor 1; M, male; Macro, macroadenoma; Micro, microadenoma; MRI, magnetic resonance imaging; PR, partial resection; p-STAT3, phosphorylated STAT3; SGA, sparsely granulated adenomas; STR, subtotal resection; T2WI, T2-weighted intensity; ULN, upper limit normal range. There are n (%) categorical data displayed. Fisher's exact test for comparison between the invasive group vs the noninvasive Group. Bold values indicate statistically significant P values. We conducted p-STAT3 immunohistochemical staining on tumor specimens from 71 patients with acromegaly and evaluated the relationship between p-STAT3 and tumor clinical features. Patients with positive p-STAT3 expression exhibited more aggressive behavior (92.00% vs 23.81%, P = 0, Table 4), larger tumor size (maximum diameter: 3.07 ± 1.56 cm vs 2.04 ± 0.93 cm; P = .007), and a higher prevalence of macroadenomas (100% vs 80.95%, P = .006). Furthermore, a high proportion of cystic degeneration (70.00% vs 23.81%, P = 0) and SG-A (40.91% vs 15.00%, P = .041) were observed in patients with positive p-STAT3 expression (Table 4). No association was observed between p-STAT3 expression and sex, age, BMI, basal hormone levels, hyperintense signals, or Ki-67 expression. Furthermore, patients with negative p-STAT3 expression exhibited a low proportion of GH levels (<1 ng/mL), with a mean of 47.62% compared with 18.00% (P = .003). Similarly, the mean IGF-1 levels were also lower in this group, at 235.81 ± 140.02 ng/mL compared with 347.10 ± 185.52 ng/mL (P = 0). In the postoperative period, a higher rate of surgical remission (52.38% vs 14.00%; P = .001, Table 4) and total resection percentage (76.19% vs 14.00%; P = 0, Table 4) were observed in patients with negative p-STAT3 expression tumors compared with those with positive p-STAT3 expression tumors.

± 185.52 ng/mL (P = 0). In the postoperative period, a higher rate of surgical remission (52.38% vs 14.00%; P = .001, Table 4) and total resection percentage (76.19% vs 14.00%; P = 0, Table 4) were observed in patients with negative p-STAT3 expression tumors compared with those with positive p-STAT3 expression tumors. Comparison of Clinical, Laboratory, Histopathological, and Imaging Factors Based on p-STAT3 Immunohistochemical Staining BMI, Body mass index; DGA, densely granulated adenoma; F, female; GH, growth hormone; GTR, gross total resection; Hyper, hyperintense; Iso, isointense; Hypo, hypointense; IGF-1, insulin-like growth factor 1; Macro, macroadenoma; M, male; Micro, microadenoma; MRI, magnetic resonance imaging; PR, partial resection; SGA, sparsely granulated adenoma; STR, subtotal resection; T2WI, T2-weighted intensity; ULN, upper limit normal range. There are n (%) categorical data displayed. Fisher's exact test for comparison between the positive group and the negative group. Bold values indicate statistically significant P values. The positive expression of p-STAT3 and the MRI were both detected using the McNemar test (P = 1), and no difference was found between them, according to the gold standard MRI. As indicated in Table 5, the Kappa coefficient test yielded a high level of agreement (κ = 0.730). A statistically significant difference was observed between the expression of SG and the “gold standard” MRI (P = 0 obtained using the McNemar test). Furthermore, the Kappa coefficient test yielded a value of 0.252, indicating a weak level of agreement (Table 6).

ient test yielded a high level of agreement (κ = 0.730). A statistically significant difference was observed between the expression of SG and the “gold standard” MRI (P = 0 obtained using the McNemar test). Furthermore, the Kappa coefficient test yielded a value of 0.252, indicating a weak level of agreement (Table 6). Comparison of Consistency Between p-STAT3 and MRI in Diagnosis of Invasive Pituitary Growth Hormone Adenoma p-STAT3, phosphorylated STAT3. When b + c = 3 + 3 < 40, McNemar's paired χ2 test was used. Fisher's exact test was used for comparison of the positive rate of 2 markers based on MRI as the gold standard for diagnosing the invasiveness of pituitary growth hormone adenoma. p-STAT3 staining was positive for invasive growth hormone adenoma. Comparison of Consistency Between Sparsely Granulated and MRI in Diagnosis of Invasive Pituitary Growth Hormone Adenoma When b + c = 0 + 18 < 40, McNemar's paired χ2 test was used. Fisher's exact test was used for comparison of the positive rate of 2 markers based on MRI as the gold standard for diagnosing the invasiveness of pituitary growth hormone adenoma. Sparsely granulated adenoma staining was used for predominant dot pattern/fibrous bodies.

When b + c = 0 + 18 < 40, McNemar's paired χ2 test was used. Fisher's exact test was used for comparison of the positive rate of 2 markers based on MRI as the gold standard for diagnosing the invasiveness of pituitary growth hormone adenoma. Sparsely granulated adenoma staining was used for predominant dot pattern/fibrous bodies. Using an ROC curve based on MRI Knosp grading, the industry standard for invasiveness, the efficacy of p-STAT3 and SG markers in diagnosing invasiveness was investigated. We found that with area under the curve values of 0.845 and 0.676, respectively, the positive expression rate of p-STAT3 was much higher than the positive expression rate of SG (Table 7 and Figure 4, P = 0), indicating that p-STAT3 differed statistically significantly in its ability to diagnose aggressiveness in GH-PAs. Comparison of Immunohistochemical Curve Analysis of 2 Markers AUC, area under the curve; p-STAT3, phosphorylated STAT3; SG, sparsely granulated. p-STAT3 staining was positive for invasive growth hormone adenoma; SG adenoma staining was used for predominant dot pattern/fibrous bodies. Receiver operating characteristic curve analysis of p-STAT3 and SG for determining invasiveness of GH-PAs based on MRI as the gold standard of invasiveness (p-STAT3 staining was positive for invasive GH-PA; SG-A staining had predominant dot pattern/fibrous bodies). GH-PA, growth hormone–secreting pituitary adenoma; p-STAT3, phosphorylated STAT3; SG-A, sparsely granular adenoma.

ern with a dot pattern of ≤8%. I, Invasive GH-PA exhibited cytoplasm that was chromophobic to pale eosinophilic (H&E). J, Noninvasive GH-PA displayed cytoplasm that was chromophobic to dense eosinophilic (H&E). GH-PA, growth hormone–secreting pituitary adenoma; H&E, hematoxylin and eosin; p-STAT3, phosphorylated STAT3. Types of cystic lesions in growth hormone–secreting pituitary adenomas. A, Presence of scattered, spotty cystic lesions; B, Some scattered, cord cystic lesions; C, Some plaque-like cystic lesions; D, Multiple cystic lesions have merged into large cystic lesions.

Immunohistochemical staining was conducted using p-STAT3/STAT3 and Keratin CAM5.2 in 71 GH-PAs. The results demonstrated that p-STAT3 was positive in 50 cases and negative in 21 cases (Figure 2 and Table 2), whereas STAT3 was positive in 69 cases and negative in only 2 cases (Figure 2 and Supplemental Digital Content 1 [http://links.lww.com/NEU/E757]). Furthermore, the CAM5.2 expression pattern identified 4 pathogenic subtypes: densely granular adenomas (n = 35; 49.29%), SG adenomas (SG-As) (n = 21; 29.58%), intermediate or transitional adenoma (n = 7; 9.82%), and Keratin-negative adenoma (n = 11.27%) (Table 2, Figure 2 and Supplemental Digital Content 2 [http://links.lww.com/NEU/E758]).22 Immunoreactivity of p-STAT3/STAT3 and CAM5.2 in Growth Hormone–Secreting Pituitary Adenomas DG-A, densely granulated adenoma; I-A, intermediate or transitional adenoma; NC-A, keratin-negative adenoma; p-STAT3, phosphorylated STAT3; SG-A, sparsely granulated adenoma.

For this purpose, the entire cohort of 71 patients was subjected to evaluation. Table 3 presents the associations between tumor invasiveness and patient characteristics, laboratory data, and imaging findings. The results demonstrated no statistically significant differences between the invasive GH-PA group and the noninvasive GH-PA group with regard to sex, age distribution, BMI, Ki-67, GH basal level, and GH level after surgery (Table 3). The maximum diameter of the tumor before surgery was significantly larger in the invasive GH-PA group (3.09 ± 1.51 cm) compared with the noninvasive GH-PA group (1.68 ± 0.77 cm; P = 0). In addition, the invasive GH-PA group exhibited a higher prevalence of macroadenoma (100% vs 80.95%; P = .006) and higher levels of IGF-1 (4 basal levels of 0.0125 ng/mL (313.50-548.80) vs 304.90 ng/mL (263.85-429.40) were associated with tumor invasiveness (P = .009). Furthermore, the group of aggressive GH-PAs demonstrated a notable increase in the prevalence of cystic variability (70.00% vs 28.57%; P = .001) and a considerable number of hyperintense signals in T2WI (44.00% vs 9.52%; P = .005) in comparison with the nonaggressive GH-PA group. As anticipated, patients with noninvasive tumors exhibited low IGF-1 levels (89.30 ng/mL [121.35-277.15] vs 352.10 ng/mL [234.40-453.40]; P = .009). In the postoperative period, a higher rate of surgical remission (38.10% vs 16.00%; P = .042, Table 3) and total resection percentage (80.95% vs 12.00%; P = 0, Table 3) were observed in patients with noninvasive tumors compared with those with invasive tumors.

We conducted p-STAT3 immunohistochemical staining on tumor specimens from 71 patients with acromegaly and evaluated the relationship between p-STAT3 and tumor clinical features. Patients with positive p-STAT3 expression exhibited more aggressive behavior (92.00% vs 23.81%, P = 0, Table 4), larger tumor size (maximum diameter: 3.07 ± 1.56 cm vs 2.04 ± 0.93 cm; P = .007), and a higher prevalence of macroadenomas (100% vs 80.95%, P = .006). Furthermore, a high proportion of cystic degeneration (70.00% vs 23.81%, P = 0) and SG-A (40.91% vs 15.00%, P = .041) were observed in patients with positive p-STAT3 expression (Table 4). No association was observed between p-STAT3 expression and sex, age, BMI, basal hormone levels, hyperintense signals, or Ki-67 expression. Furthermore, patients with negative p-STAT3 expression exhibited a low proportion of GH levels (<1 ng/mL), with a mean of 47.62% compared with 18.00% (P = .003). Similarly, the mean IGF-1 levels were also lower in this group, at 235.81 ± 140.02 ng/mL compared with 347.10 ± 185.52 ng/mL (P = 0). In the postoperative period, a higher rate of surgical remission (52.38% vs 14.00%; P = .001, Table 4) and total resection percentage (76.19% vs 14.00%; P = 0, Table 4) were observed in patients with negative p-STAT3 expression tumors compared with those with positive p-STAT3 expression tumors. Comparison of Clinical, Laboratory, Histopathological, and Imaging Factors Based on p-STAT3 Immunohistochemical Staining

We conducted p-STAT3 immunohistochemical staining on tumor specimens from 71 patients with acromegaly and evaluated the relationship between p-STAT3 and tumor clinical features. Patients with positive p-STAT3 expression exhibited more aggressive behavior (92.00% vs 23.81%, P = 0, Table 4), larger tumor size (maximum diameter: 3.07 ± 1.56 cm vs 2.04 ± 0.93 cm; P = .007), and a higher prevalence of macroadenomas (100% vs 80.95%, P = .006). Furthermore, a high proportion of cystic degeneration (70.00% vs 23.81%, P = 0) and SG-A (40.91% vs 15.00%, P = .041) were observed in patients with positive p-STAT3 expression (Table 4). No association was observed between p-STAT3 expression and sex, age, BMI, basal hormone levels, hyperintense signals, or Ki-67 expression. Furthermore, patients with negative p-STAT3 expression exhibited a low proportion of GH levels (<1 ng/mL), with a mean of 47.62% compared with 18.00% (P = .003). Similarly, the mean IGF-1 levels were also lower in this group, at 235.81 ± 140.02 ng/mL compared with 347.10 ± 185.52 ng/mL (P = 0). In the postoperative period, a higher rate of surgical remission (52.38% vs 14.00%; P = .001, Table 4) and total resection percentage (76.19% vs 14.00%; P = 0, Table 4) were observed in patients with negative p-STAT3 expression tumors compared with those with positive p-STAT3 expression tumors. Comparison of Clinical, Laboratory, Histopathological, and Imaging Factors Based on p-STAT3 Immunohistochemical Staining BMI, Body mass index; DGA, densely granulated adenoma; F, female; GH, growth hormone; GTR, gross total resection; Hyper, hyperintense; Iso, isointense; Hypo, hypointense; IGF-1, insulin-like growth factor 1; Macro, macroadenoma; M, male; Micro, microadenoma; MRI, magnetic resonance imaging; PR, partial resection; SGA, sparsely granulated adenoma; STR, subtotal resection; T2WI, T2-weighted intensity; ULN, upper limit normal range.

Hyper, hyperintense; Iso, isointense; Hypo, hypointense; IGF-1, insulin-like growth factor 1; Macro, macroadenoma; M, male; Micro, microadenoma; MRI, magnetic resonance imaging; PR, partial resection; SGA, sparsely granulated adenoma; STR, subtotal resection; T2WI, T2-weighted intensity; ULN, upper limit normal range. There are n (%) categorical data displayed. Fisher's exact test for comparison between the positive group and the negative group. Bold values indicate statistically significant P values.

The positive expression of p-STAT3 and the MRI were both detected using the McNemar test (P = 1), and no difference was found between them, according to the gold standard MRI. As indicated in Table 5, the Kappa coefficient test yielded a high level of agreement (κ = 0.730). A statistically significant difference was observed between the expression of SG and the “gold standard” MRI (P = 0 obtained using the McNemar test). Furthermore, the Kappa coefficient test yielded a value of 0.252, indicating a weak level of agreement (Table 6). Comparison of Consistency Between p-STAT3 and MRI in Diagnosis of Invasive Pituitary Growth Hormone Adenoma p-STAT3, phosphorylated STAT3. When b + c = 3 + 3 < 40, McNemar's paired χ2 test was used. Fisher's exact test was used for comparison of the positive rate of 2 markers based on MRI as the gold standard for diagnosing the invasiveness of pituitary growth hormone adenoma. p-STAT3 staining was positive for invasive growth hormone adenoma. Comparison of Consistency Between Sparsely Granulated and MRI in Diagnosis of Invasive Pituitary Growth Hormone Adenoma When b + c = 0 + 18 < 40, McNemar's paired χ2 test was used. Fisher's exact test was used for comparison of the positive rate of 2 markers based on MRI as the gold standard for diagnosing the invasiveness of pituitary growth hormone adenoma. Sparsely granulated adenoma staining was used for predominant dot pattern/fibrous bodies.

We do not know whether GH-PA groups express p-STAT3/STAT3 differently. We included 71 patients (1:1.45 male to female ratio) to evaluate p-STAT3/STAT3 expression between the 2 groups. The invasive GH-PA group had high STAT3 expression, but there was no difference between the 2 groups. Invasive GH-PAs had more p-STAT3 than noninvasive ones. p-STAT3–positive GH-PAs had lower remission and resection rates. We investigated whether p-STAT3 could diagnose invasiveness in GH-PAs more accurately than SG. Cystic invasive GH-PAs had more p-STAT3–positive expression. In a previous study, the expression of STAT3 was assessed at the molecular level to compare the number of GH-PAs with nonsecreting PAs.11The immunohistochemical study showed significant expression of STAT3 in GH-PAs (100%), which is consistent with this observation. Several recent studies have shown that p-STAT3 expression levels are significantly increased in cancer tissues compared with control tissues.23,24 Furthermore, there is a clear association between p-STAT3 and tumor aggressiveness. In addition, numerous previous meta-analyses have suggested that the level of p-STAT3 expression could be used as a novel clinicopathological biomarker to predict poor prognosis in various types of human malignancies.25-30 Our study found a positive correlation between higher levels of p-STAT3 expression and tumor grade and invasion.31

rous previous meta-analyses have suggested that the level of p-STAT3 expression could be used as a novel clinicopathological biomarker to predict poor prognosis in various types of human malignancies.25-30 Our study found a positive correlation between higher levels of p-STAT3 expression and tumor grade and invasion.31 Our results showed that p-STAT3 expression did not correlate with Ki-67 expression. However, our study found a positive correlation between higher levels of p-STAT3 expression and preoperative IGH-1 levels, preoperative T2 signaling, SG-A, and tumor grade and invasion. It is controversial to use keratin as a subtype to determine the postoperative remission rate of GH adenoma.14,22,32 Our study found that other factors, such as remission and resection rates, were linked to p-STAT3 expression. P-STAT3 expression may be used to predict the outcome of invasive GH-PA. This could help with further research into invasive GH-PA. Studies have shown that STAT3 and its downstream target genes, including MMP9, could be activated by IL-6/IL-6R or IGF1/IGF1R.33-35 We speculated that STAT3 promotes GH adenoma invasion because IL-6/IL-6R or IGF1/IGF1R combines to activate STAT3 and leads to downstream gene expression, thereby promoting the expression of MMP9 and leading to increased invasiveness.

am target genes, including MMP9, could be activated by IL-6/IL-6R or IGF1/IGF1R.33-35 We speculated that STAT3 promotes GH adenoma invasion because IL-6/IL-6R or IGF1/IGF1R combines to activate STAT3 and leads to downstream gene expression, thereby promoting the expression of MMP9 and leading to increased invasiveness. We performed immunohistochemical staining for p-STAT3 and CAM5.2 in 71 patients each to evaluate the efficacy of SG and p-STAT3 in diagnosing the invasiveness of GH-PAs. The frequency distribution of histological subtypes in our sample is consistent with those reported in the literature, with 49.29% representing densely granular adenomas, 29.58% representing SG-A, 9.86% representing intermediate or transitional adenoma, and 11.27% representing Keratin-negative adenoma22 (Table 2). This study showed that p-STAT3 was more useful than SG for diagnosing the invasiveness of GH-PAs. MRI was used to determine the invasiveness of PAs, and p-STAT3 was more accurate than SG in diagnosing the aggressiveness of GH-PAs. This was confirmed by ROC curve analysis, which showed that p-STAT3 was more accurate than SG. We suggested that p-STAT3 could be used to indicate invasiveness in GH-PAs. We will use a larger sample size in future studies to validate this finding.

T3 was more accurate than SG in diagnosing the aggressiveness of GH-PAs. This was confirmed by ROC curve analysis, which showed that p-STAT3 was more accurate than SG. We suggested that p-STAT3 could be used to indicate invasiveness in GH-PAs. We will use a larger sample size in future studies to validate this finding. Our study shows for the first time that cystic aggressive GH-PAs are positive for p-STAT3. There are 2 main theories about how cystic changes develop in PAs. One is that bleeding causes cystic changes. The other is that lack of blood supply and tissue death lead to cystic transformation.36-38 Previous studies have shown that GH-PAs are hypermetabolic tumors.39,40 In aggressive GH-PAs, cystic degeneration results from hypermetabolic growth causing hypoxia and ischemia in the tumor tissue.41 Research has shown a correlation between cystosis of pituitary tumors and the hypermetabolic status of the tumors.42 This hypermetabolic status has been found to be associated with tumor aggressiveness.43

GH-PAs, cystic degeneration results from hypermetabolic growth causing hypoxia and ischemia in the tumor tissue.41 Research has shown a correlation between cystosis of pituitary tumors and the hypermetabolic status of the tumors.42 This hypermetabolic status has been found to be associated with tumor aggressiveness.43 The results of our study showed that invasive GH-PAs had greater cystic variability, whereas invasive cystic GH-PAs were predominantly large adenomas, which is consistent with previous research.38,44,45 The Wallenberg effect describes the observation that tumor cells have a higher metabolic state, deriving more energy from glycolysis than normal cells. This metabolic difference is associated with tumor aggressiveness.46-48 We reasoned that inhibiting the high metabolic state of GH-PAs may be a new therapeutic approach to treat the invasiveness of GH-PAs while reducing the stroke rate of GH-PAs. An important limitation of this study was the small number of patients, mainly because several samples were excluded because of inadequacy. In GH-PAs, only p-STAT3/STAT3 expression was assessed, so important data such as MMP2/MMP9 were not available. Furthermore, immunohistochemical staining for p53 was not performed on the tumor specimens. Our study did not look at how p-STAT3 affects the invasiveness of GH-PAs in vitro. Further research is needed to understand the link between p-STAT3 and the aggressiveness of GH-PAs. More research is needed to confirm our findings. This could involve large clinical trials in different places.

An important limitation of this study was the small number of patients, mainly because several samples were excluded because of inadequacy. In GH-PAs, only p-STAT3/STAT3 expression was assessed, so important data such as MMP2/MMP9 were not available. Furthermore, immunohistochemical staining for p53 was not performed on the tumor specimens. Our study did not look at how p-STAT3 affects the invasiveness of GH-PAs in vitro. Further research is needed to understand the link between p-STAT3 and the aggressiveness of GH-PAs. More research is needed to confirm our findings. This could involve large clinical trials in different places.

Our results show that p-STAT3 is a better marker for predicting the aggressiveness of GH-PA than SG. P-STAT3 could be used to predict how aggressive GH-PA is. Cystic invasive GH-PA had more p-STAT3–positive expression. Understanding the pathways of GH-PA invasiveness could lead to new treatments.