TYPE Original Research PUBLISHED 16 September 2022 DOI 10.3389/fendo.2022.984877 Factors associated with therapeutic response in OPEN ACCESS EDITED BY acromegaly diagnosed in the Lucio Vilar, Federal University of Pernambuco, Brazil elderly in Spain REVIEWED BY Adriana G. Ioachimescu, Betina Biagetti 1*, Pedro Iglesias 2, Rocı́o Villar-Taibo 3, Emory University, United States Andrea Lania, Marı́a-Dolores Moure 4,5, Miguel Paja 6, Marta Araujo-Castro 7, Humanitas University, Italy Jessica Ares 8,9, Cristina Álvarez-Escola 10, Almudena Vicente 11, *CORRESPONDENCE Betina Biagetti Èlia Álvarez Guivernau 12,13, Iria Novoa-Testa 14, bbiagetti@vhebron.net Fernando Guerrero Perez 15, Rosa Cámara 16, Manel Puig-Domingo mpuigd@igtp.cat Beatriz Lecumberri 10, Carlos Garcı́a Gómez 2, SPECIALTY SECTION Ignacio Bernabéu 3, Laura Manjón 8,9, Sonia Gaztambide 4,17, This article was submitted to Pituitary Endocrinology, Fernando Cordido 14, Susan M. Webb 12,13, a section of the journal Edelmiro Luis Menéndez-Torre 8,9, Juan J. Dı́ez 2,18, Frontiers in Endocrinology Rafael Simó 1 and Manel Puig-Domingo 19* RECEIVED 02 July 2022 ACCEPTED 30 August 2022 1 Endocrinology Department, Diabetes and Metabolism Research Unit, Vall d’Hebron University PUBLISHED 16 September 2022 Hospital and Vall d’Hebron Research Institute (VHIR), Universidad Autónoma de Barcelona, Barcelona, Spain, 2 Department of Endocrinology, Hospital Universitario Puerta de Hierro CITATION Majadahonda, Madrid, Spain, 3 Endocrinology Department, University Hospital of Santiago de Biagetti B, Iglesias P, Villar-Taibo R, Compostela, Neoplasia and Differentiation of Endocrine Cells Group, Instituto de Investigación Moure M-D, Paja M, Araujo-Castro M, Sanitaria-IDIS (Health Research Institute), Santiago de Compostela, Spain, 4 Endocrinology Ares J, Álvarez-Escola C, Vicente A, Department, Cruces University Hospital, Biocruces Bizkaia, Endo-European Reference Networks Álvarez Guivernau È, Novoa-Testa I, (ERN), Barakaldo, Spain, 5 Neoplasia and Differentiation of Endocrine Cells Group, Instituto de Guerrero Perez F, Cámara R, Investigación Sanitaria-IDIS (Health Research Institute), Santiago de Compostela, Spain, Lecumberri B, Garcı´a Gómez C, 6 Department of Endocrinology, Hospital Universitario Basurto, Bilbao, Spain, 7 Department of Bernabéu I, Manjón L, Gaztambide S, Endocrinology and Nutrition, Hospital Universitario Ramón y Cajal, Madrid, Spain, 8 Department of Cordido F, Webb SM, Menéndez- Torre EL, Dı´ez JJ, Simó R and Puig- Endocrinology and Nutrition, Hospital Universitario Central de Asturias, Asturias, Spain, 9 Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), Centro de Investigación Biomédica en Red Domingo M (2022) Factors associated Enfermedades Raras (CIBERER), Universidad de Oviedo, Oviedo, Spain, 10 Department of with therapeutic response in Endocrinology, Hospital Universitario La Paz, Madrid, Spain, 11 Department of Endocrinology and acromegaly diagnosed in the elderly in Spain. Nutrition, Hospital Universitario de Toledo, Toledo, Spain, 12 Department of Endocrinology, Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER), Hospital Sant Pau, Institut Front. Endocrinol. 13:984877. d’Investigacions Biomèdiques (IIB) SPau, Barcelona, Spain, 13 Department of Medicine, Universitat doi: 10.3389/fendo.2022.984877 Autónoma de Barcelona, Barcelona, Spain, 14 Endocrinology and Nutrition Department, A Coruña University Hospital and A Coruña University, A Coruña, Spain, 15 Department of Endocrinology, Hospital Universitari de Bellvitge, L’Hospitalet de Llobregat, Barcelona, Spain, 16 Endocrinology and Nutrition Service, La Fe University Hospital, Valencia, Spain, 17 Health Research Institute, Universidad País Vasco - Euskal Herriko Unibertsitatea (UPV-EHU), Centro de Investigación Biomédica en Red Enfermedades Raras (CIBERER), Centro de Investigación Biomédica en Red Diabetes y Enfermedades Metabólicas (CIBERDEM), Barakaldo, Spain, 18 Department of Medicine, Instituto de Investigación Sanitaria Puerta de Hierro Segovia de Arana, Universidad Autónoma de Madrid, Majadahonda, Spain, 19 Endocrinology and Nutrition Service, Germans Trias Hospital and Research Institute, Autonomous University of Barcelona, Badalona, Spain Context: Some reports suggest that acromegaly in elderly patients has a more benign clinical behavior and could have a better response to first-generation long-acting somatostatin receptor ligands (SRL). However, there is no specific therapeutic protocol for this special subgroup of patients. Frontiers in Endocrinology 01 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 Objective: This study aimed at identifying predictors of response to SRL in elderly patients. Design: Multicentric retrospective nationwide study of patients diagnosed with acromegaly at or over the age of 65 years. Results: One-hundred and eighteen patients (34 men, 84 women, mean age at diagnosis 71.7 ± 5.4 years old) were included. Basal insulin-like growth factor type 1 (IGF-1) above the upper limit of normal (ULN) and growth hormone (GH) levels (mean ± SD) were 2.7 ± 1.4 and 11.0 ± 11.9 ng/ml, respectively. The mean maximal tumor diameter was 12.3 ± 6.4 mm, and up to 68.6% were macroadenoma. Seventy-two out of 118 patients (61.0%) underwent surgery as primary treatment. One-third of patients required first-line medical treatment due to a rejection of surgical treatment or non-suitability because of high surgical risk. After first-line surgery, 45/72 (63.9%) were in disease remission, and 16/34 (46.7%) of those treated with SRL had controlled disease. Patients with basal GH at diagnosis ≤6 ng/ml had lower IGF-1 levels and had smaller tumors, and more patients in this group reached control with SRL (72.7% vs. 33.3%; p < 0.04) [OR: 21.3, IC: 95% (2.4–91.1)], while male patients had a worse response [OR: 0.09, IC 95% (0.01–0.75)]. The predictive model curve obtained for SRL response showed an AUC of 0.82 CI (0.71–0.94). Conclusions: The most frequent phenotype in newly diagnosed acromegaly in the elderly includes small adenomas and moderately high IGF-1 levels. GH at diagnosis ≤6 ng/ml and female gender, but not age per se, were associated with a greater chance of response to SRL. KEYWORDS acromegaly, elderly, growth hormone, IGF-1 (insulin-like growth factor 1), somatostatin analog, surgery, Spain Introduction (15). In fact, in the group of very old subjects (arbitrarily defined as those older than 75 years), the rate of patients’ rejection to The behavior of GH-secreting tumors is heterogeneous and surgery could be high, although data in this regard are scarce. differs between patients (1–3). Although some evidence suggests In order to gain further insights into the treatment options that acromegaly could be less aggressive when diagnosed in and patients’ response when acromegaly is diagnosed in elderly elderly patients (4, 5), there are no age-adapted protocols and subjects, we have examined these questions in the real-world transsphenoidal surgery is also recommended as primary setting, performing a multicentric, nationwide, observational treatment in those patients (6–9). In addition, older patients study of patients diagnosed with acromegaly at or over the age are underrepresented in clinical trials due to eligibility criteria, of 65 years. resulting in scarce data on outcomes in this subset of the population (10–12). However, probably due to the increase in life expectancy, acromegaly is diagnosed more frequently than Patients and methods previously in elderly people and it is not expected to decrease (13, 14). This observational retrospective, multicentric, nationwide Elderly patients with acromegaly are also a heterogeneous study ranged from 1995 to 2020 and included patients newly group because some of them suffer from frailty and other age- diagnosed with acromegaly at or over the age of 65 years. The dependent diseases such as cognitive impairment, which can add study was endorsed by the Spanish Society of Endocrinology and an additional challenge when deciding on treatment options Nutrition (SEEN) and was disseminated to all members of the Frontiers in Endocrinology 02 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 Neuroendocrinology Task Force of the SEEN, which includes was defined when those patients under medical treatment most of the endocrinologists who take care of acromegaly presented IGF-1 levels within the specific age- and sex- patients in Spain. adjusted reference range or <1.2 ULN. Cured or acromegaly in A specific registry was set up to collect real-life clinical remission was defined when the IGF-1 level was normal without features of acromegaly in the elderly and usual patients’ treatment, for more than 6 months. management modalities as well as the factors related to Hormonal status was evaluated at diagnosis, after any therapeutic response. A final version of the protocol was treatment, and at the last visit. approved as a multicenter observational study with drugs by the Ethics Committees of the Vall d’Hebron University Hospital (number: PR(AG)318/2021). The study was conducted according to the mandates of the Declaration of Helsinki and Statistical analysis good clinical practices. The patients’ confidential information A descriptive analysis was performed. Numerical data are was protected according to the Spanish data protection expressed as means and standard deviation (SD) (Gaussian national law. distribution) or as medians (p50) and interquartile ranges (IQR) (non-Gaussian distribution). Categorical data are described using numbers and percentages. A comparison Inclusion and exclusion criteria between two groups was performed using Student’s t-test or Mann–Whitney U-test for numerical data and chi-square test/ Inclusion criteria were all patients with pituitary acromegaly Fisher exact test for categorical variables as appropriate. Median diagnosed at or above the age of 65 years attending the different values across more than 2 groups [i.e., surgery vs. somatostatin participating hospitals with at least 1 year of follow-up at the receptor ligand (SRL) vs. dopamine agonist] were compared same hospital. Patients with incomplete data were excluded from using the Kruskal–Wallis test with Bonferroni adjustment. We the study. investigated which variables influence the response to SRL; predictors with p < 0.20 after Spearman’s rank test were included in multiple logistic regression, and a stepwise Variables and measurements backward selection approach was used to identify the best predictive model. The model fit and calibration were assessed Clinical data, hormonal workup, imaging procedures, by Akaike’s and Bayesian information criterion, and comorbidities, therapeutic modalities, and outcomes were the Hosmer–Lemeshow goodness-of-fit test, and the recorded. For each patient, the following parameters were discrimination power by calculating the area under the analyzed: year and age of diagnosis, gender, body mass index receiver operating characteristic (ROC) curves (AUC). Finally, (BMI), specific clinical manifestations and comorbidities related we formulated a predictive score useful for therapeutic decisions to acromegaly, the estimated duration of symptoms before the in clinical practice. Results were expressed as odds ratio (OR) diagnosis of acromegaly, and growth hormone (GH) and and 95% CI. A two-sided P-value of <0.05 was considered insulin-like growth factor type 1 (IGF- 1) levels. Regarding significant. Statistical analyses were performed using the tumor features, sinus invasion evaluated by Knosp STATA16 statistical package (USA) for Windows. classification, sellar involvement, and chiasma compression were evaluated. Finally, we also evaluated first- and second- line therapies used for acromegaly, outcomes on efficacy and side effects of surgical and medical treatments, and disease activity at Results the last available visit of follow-up. Patients’ characteristics: Biochemical and tumor features Diagnosis of acromegaly and disease control Fifteen centers participated in the registry including patients diagnosed in the study period (1995–2020). From a total of 1,069 The diagnosis of acromegaly was made according to the patients with acromegaly, 126 (11.8%) were older than 65 years clinical practice guideline criteria in force at the time of at diagnosis. Among these 126 patients, two had an ectopic diagnosis. Incidental diagnosis was established when the acromegaly and six were not included in the analyses due to diagnosis of acromegaly was made as a result of a cerebral incomplete data; therefore finally, 118 patients were analyzed. MRI performed for any reason, which detected a pituitary Of the total, 84 (71.2%) were women and 34 (28.8%) men. adenoma, leading to the acromegaly diagnosis. Active disease The mean age at diagnosis was 71.7 ± 5.4 years without was defined as IGF-1 level above 1.2 ULN. Controlled disease differences by gender. Patients with acromegaly diagnosed Frontiers in Endocrinology 03 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 beyond the age of 75 years were also more frequently women There was no postoperative mortality. Surgical (34.5% vs. 14.7%, P = 0.02) (Table 1). complications were present in 25/63 (39.7%) of registered The median time with specific symptoms before diagnosis cases, with nine diabetes insipidus (four transient), three (SBD) was 48.0 (100) [median (IQR)] months. In 20 (16.9%) cerebrospinal fluid leakage, one III cranial nerve palsy, one patients, the diagnosis was incidental and the median time with epistaxis, two transient syndrome of inappropriate antidiuretic SBD in this group was shorter [10 (85.0) months vs. 48 (100) hormone secretion, and 17 new cases of hypopituitarism, and (P < 0.02)] compared with patients in which the diagnosis was four cases had some subjective neurocognitive impairment. non-incidental. The maximal tumor diameter with accurate Regarding medical treatment, 13 out of 34 patients (38.2%) radiological tumor invasion description was available in 105 who were under first-line SRL experienced side effects: nine patients, and it was 12.3 ± 6.4 mm. GH and IGF-1 ULN were digestive, five mild hyperglycemia, two muscle pain, and four 11.0 (11.9) ng/ml and 2.7 ± 1.4 ULN, respectively, without fatigue. However, in only three cases, the treatment had to be gender differences (Table 1). Despite having relatively small stopped due to side effects. tumors (the maximum diameter was 12.3 ± 6.4), 68.6% of patients had a macroadenoma. Regarding the presence of acromegaly comorbidities, the Disease control and follow-up most reported were hypertension in 91 cases (77.1%) and arthropathy in 54 (46.2%). Type 2 diabetes was only present in After first-line surgery, 45/72 (63.9%) patients were in 40 (33.9%), which was not higher than the prevalence reported disease remission, and 16/34 (46.7%) of those treated with for acromegaly populations of younger age (16). first-line SRL had controlled disease, without statistical differences between both treatment modalities. In the follow-up, a second-line treatment was required in 57 First-line treatment options, (48.3%) patients, of which 28, 21, and seven received first-line complications, and side effects treatment, surgery, SRL, or DA respectively. The most frequent treatment used as second-line was SRL alone or associated with In total, 72 out of 118 (61.0%) underwent surgery. In 15 DA (Figure 1). Radiotherapy was also indicated in six patients, (12.7%) patients, surgery was contraindicated by the anesthetist four after surgery, and two after SRL. or the pituitary care unit team, due to the high surgical risk for The median of follow-up was 8.6 years [median (IQR) 103, age and associated comorbidities, and 27 of the remaining 103 (72.3) months]. At their last visit, 41/118 (34.8%) patients were (26.2%) patients refused surgery. A total of 34 patients (28.9%) in remission, 61/118 (51.7%) had controlled disease on medical were treated with first-line SRL (50% with lanreotide and 50% treatment, and the remaining 16/118 (13.6%) had active non- with octreotide) and nine (7.6%) with dopamine analogues controlled disease, five of them without medical treatment (three (DA), all of them with cabergoline (Figure 1); three patients refused treatment after diagnosis and the other two refused any refused any kind of treatment. kind of treatment after experiencing side effects after medical TABLE 1 Baseline characteristics at diagnosis of acromegaly. All Women Men P-value N (%) 118 84 (71.2) 34 (21.8) Age at diagnosis, years (mean ± SD) 71.7 ± 5.4 72.3 ± 5.4 70.4 ± 5.2 0.09 Older than 75 years old n (%) 34 (28.8) 29 (34.5) 5 (14.7) 0.02 Older than 80 years old n (%) 9 (7.6%) 6 (7.1) 3 (8.8) 0.50 SBD, months (median, IQR)a 48.0 (100.0) 49.0 (101.0) 24.0 (96.0) 0.29 Incidental n (%) 20 (16.9) 12 (14.3) 8 (23.5) 0.29 GH, ng/mL (median, IQR)a 11.0 (11.9) 7.8 (11.2) 4.2 (7.2) 0.14 IGF-1, ng/mL (mean ± -SD) 610.1 ± 63.0 621.4 ± 60.3 582.9 ± 68.5 0.53 IGF-1 (ULN) 2.7 (1.4) 2.9 (1.5) 2.5 (1.0) 0.15 Diameter max, mm (mean ± -SD)b 12.3 ± 6.4 12.0 ± 6.6 12.7 ± 7.1 0.83 Macroadenoma n (%) 81 (68.6) 59 (70.2) 22 (64.7) 0.47 Extrasellar n (%)b 50 (47.6) 35 (48.0) 15 (46.9) 0.97 Chiasm compression n (%) 15 (14.3) 8 (11.0) 7 (21.9) 0.13 Sinus invasion n (%)b 49 (46.7) 34 (47.2) 15 (46.9) 0.71 SD, standard deviation; IQR, interquartile range; SBD, symptoms before diagnosis; GH, growth hormone; IGF-1, insulin-like growth factor 1; max, maximum. aReported in 110 patients. b Reported in 105 patients. Frontiers in Endocrinology 04 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 FIGURE 1 Treatment options. First-line on the left and second-line on the right. Results are expressed in “n” and percentage. WaS (wait and see) SRL, first generation somatostatin receptor ligand; DA, dopamine agonist ( all cabergoline); PEG, pegvisomant; PAS, pasireotide; RT, radiotherapy. therapy; all of them were patients older than 70). In those 60.0 months; P < 0.01). Incidental diagnosis showed a statistical patients receiving active treatment, IGF-1 concentrations trend in those older >75 [9/34 (26.5%) vs. 11/84 (13.1%), markedly decreased from baseline to the last visit, without P = 0.07; respectively]. Neither GH, IGF-1 ULN, tumor size, gender differences (Figure 2). tumor invasion, nor percentage of comorbidities at diagnosis showed statistical differences when compared to those cases diagnosed between 65 and 75 years of age. Of these 34 patients Disease characteristics in patients of or aged 75 or older at diagnosis, only 12 (35.3%) underwent above the age of 75 years surgery, while 21 (61.8%) received medical treatment [16 (47.1%) SRL and five (14.7%) DA]. No statistical differences A total of 34 out of 118 (28.8%) were patients ≥75 years old were found in the percentage of patients with controlled disease (Table 2); male gender was less frequently represented in these between both groups. older patients [5 (14.7%) vs. 29 (85.3%); P < 0.02]. The median time with referred SBD was shorter in the older group (24.0 vs. Clinical characteristics and therapeutic response according to basal GH levels We evaluated if basal GH levels were related to disease control after SRL. For this, we assessed the optimal GH cutoff point with an ROC analysis. The binomial area under the receiver operating characteristic curve (AUC) with this unique variable was 0.72 (IC: 0.57–0.86), and the optimal cutoff was 6 ng/ml (sensibility: 85.20%, specificity: 63.80%, correct classification: 67.20%). Thus, we further divided the sample in patients according to GH levels. Table 3 shows, age, gender, and tumor characteristics of both groups. The GH ≤6-ng/ml group included 49/110 (44.5%) patients; male gender was more frequent in this group [20 (40.8%) vs. 13 (21.3%) P < 0.02], and the time in months with referred SBD was shorter (24.0 vs. 60.0 months; P < 0.01), although incidental diagnosis was not FIGURE 2 more frequent. Regarding biologic and tumor features, the GH IGF-1 x ULN at baseline and in the last follow-up according gender. ≤6-ng/ml group had lower IGF-1 levels, as well as smaller and less extrasellar tumors (Table 3). The frequency of comorbidities Frontiers in Endocrinology 05 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 TABLE 2 Differences between patients older or younger than 75 years. ≥75 years 65–75 years P-value N (%) 118 34 (28.8) 84 (71.2) Male gender 34 (28.8) 5 (14.7) 29 (34.5) 0.02 SBD, months (median, IQR)a 48.0 (100) 24.0 (50.0) 60.0 (99.0) 0.01 Incidental (n, %) 20 (16.9) 9 (26.5) 11 (13.1) 0.07 GH, ng/mL (median, IQR)a 11.0 (11.9) 12 (12.0) 10 (11.9) 0.50 IGF-1, ng/mL (mean ±- SD) 610.1 (63.0) 582.9 (68.5) 621.4 (60.3) 0.53 IGF-1 (ULN) 2.7 (1.4) 2.5 (1.0) 2.9 (1.5) 0.15 Diameter max, mm (mean ±- SD)a 12.3 ± 6.4 12.9 ± 6.4 12.4 ± 6.5 0.84 Macroadenoma n (%) 81 (68.6) 22 (64.7) 59 (70.2) 0.35 Extrasellar n (%)b 50 (47.6) 11 (36.7) 39 (52.0) 0.12 Chiasm compression n (%) 15 (14.3) 3 (10.0) 12 (15.8) 0.33 b Sinus invasion n (%) 49 (46.7) 16 (52.3) 33 (42.92) 0.22 a SD, standard deviation; IQR, interquartile range; SBD, symptoms before diagnosis; GH, growth hormone; IGF-1, insulin-like growth factor 1; max, maximum. Reported in 110 patients. b Reported in 105 patients. at diagnosis was similar between the two groups. In the group and BMI. The model was well-fitted according to Akaike’s and with GH ≤6 ng/ml, more patients achieved disease control with Bayesian information criterion, and the likelihood ratio SRL (72.7% vs. 33.3%; P < 0.04) and showed a statistical trend to compared with the null model was <0.01. The binomial AUC a higher degree of surgical cure (73.3% vs. 52.8%; P = 0.08) was 0.82 CI (0.71–0.94) (Figure 4 and Supplementary Material). (Table 3 and Figure 3). The odds ratio (OR) obtained for GH ≤6 ng/ml was 21.3 [IC: 95% (2.4–91.1)]. Thus, patients with GH ≤6 ng/ml were 21.3 times more likely to respond to SRL than those with GH >6 ng/ Predictive model of response to SRL ml, adjusted by (basal ULN_IGF-1, gender, maximum tumor diameter, and BMI). Male gender had a worse response with an We performed a stepwise backward selection approach to OR of 0.09 [IC 95%:(0.01–0.75)], while the OR of SRL response identify the best predictive model for response to SRL. The for women was 11.11 [IC 95% (1.33–100)]. included variables were age at diagnosis, gender, basal GH, basal After log regression, with the predict command of the ULN_IGF-1, BMI, maximum tumor diameter, extrasellar STATA package, we generated a new variable called extension, and type 2 diabetes. The reduced and parsimonious “predictions” that estimated the probability of response to SRL model that best explained SRL response included five predictors: in our sample. Then, a predictive table was obtained. For this, basal GH, basal ULN_IGF-1, gender, maximum tumor diameter, continuous variables were transformed into categorical variables TABLE 3 Clinical characteristics and therapeutic response according to GH levels. All GH ≤6 ng/dl GH >6 ng/dl P-value N (%) 110 49 61 Age (mean ±- SD) 71.6 ± 5.25 72.1 ± 5.8 71.4 ± 5.0 0.54 Male gender n (%) 33 (30.0) 20 (40.8) 13(21.3) 0.02 BMI kg/m2 (mean ±- SD) 29.3 ± 4.9 30.9 ± 5.6 28.0 ± 3.9 0.01 SBD, months (median, IQR)a 48 (100) 24.0 (72.0) 60.0 (96.0) 0.01 a Incidental n (%) 20 (18.2) 7 (14.3) 13 (22.0) 0.33 IGF-1, ng/mL (mean ± SD) 586.2 ± 242.8 427.8 ± 170.8 674.0 ± 230.8 0.01 IGF-1 (ULN) 2.6 ± 1.1 1.9 ± 0.8 3.0 ± 1.1 0.01 Diam max, mm (mean ±- SD)a 12.3 ± 5.4 10.4 ± 4.2 13.3 ± 6.1 0.01 Macroadenoma 75 (68.2) 27 (55.1) 48 (78.7) 0.01 Extrasellar n (%)b 47 (44.7) 11 (28.2) 36 (50.7) 0.01 Surgery_remission n (%) 41/66 (62.1) 22/30 (77.3) 19/36 (52.8) 0.08 Controlled with SRL n (%) 15/32 (46.9) 8/11 (72.7) 7/21 (33.3) 0.04 SD, standard deviation; SBD, symptoms before diagnosis; GH, growth hormone; IGF-1, insulin-like growth factor 1; max, maximum. aReported in 110 patients. bReported in 105 patients. Frontiers in Endocrinology 06 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 FIGURE 3 FIGURE 4 Percentage of patients controlled after surgery or SRL, according ROC curve for SRL response prediction performed according to basal GH levels. the selected model. (Supplementary Material). The results of the predictive table are MRI. However, the prevalence of incidental diagnosis in our shown in Table 4. series was virtually the same at 17% than the percentage recently reported by Giraldi et al. (21) in younger operated patients. The disease characteristics of acromegaly diagnosed at older Discussion ages present some differences with those reported in the registries and large series (16) in younger patients, as Acromegaly is a chronic disease, which is diagnosed at any exemplified in the present study. In our cohort, although up to age, even in the elderly population. This is the first Spanish series 68% of the patients had a macroadenoma, the median of comprising patients with acromegaly diagnosed at or over the maximum tumor size was smaller than that reported in age of 65 years. This subset represents up to 11.8% of the younger cases, with a smaller proportion consisting of invasive diagnoses of acromegaly in the present cohort in line with pituitary adenomas. These findings support the concept that in other registries (10, 11, 17–19). The median of time with SBD the elderly, GH-secreting tumors are less aggressive. However, was 4 years (48.0 months), similar but slightly reduced than the biochemical data at diagnosis were similar to those previously one reported in the general Acromegaly Spanish Registry (20). reported in the Spanish Acromegaly Registry (22). This supports the idea that the reported increased incidence of Therapeutic patterns in our series did not differ from general acromegaly diagnosis in the elderly (11, 13, 14) is not a delay in recommendation from guidelines, although in a subset of the diagnosis. Probably, it may be the consequence of an increase patients, surgery was contraindicated and in up to 26.2% in life expectancy, particularly in the Spanish population, surgery was refused, although postoperative mortality or although it could also be associated with a better awareness of serious peri-surgical morbidity was not high. Also, the the disease and in part with some degree of enhanced incidental percentage of patients controlled with first-line surgery was diagnosis over recent years due to the extensive use of CT and 63.9%, which did not differ from those controlled after first- TABLE 4 Estimated probability of response to SRL expressed in percentage. Gender and IGF (× ULN) Female Male IGF-1<1.5 × IGF-11.5-2.5 × IGF-1>2.5 × IGF-1<1.5 × IGF-11.5-2.5 × IGF-1>2.5 × ULN ULN ULN ULN ULN ULN GH >6 ng/mL 42.7 29.1 12.1 7.8 1.3 4.3 GH ¾6 ng/mL 92.0 86.1 67.0 56.7 43.0 23.0 SRL, somatostatin receptor ligand. The table shows the estimated probability of response to SRL, expressed in percentage according gender IGF-1 and GH baseline levels. Frontiers in Endocrinology 07 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 line SRL (46.7%) and was not particularly different to what has se and not age are what determine clinical behavior and been described for young adults (23). Four previous studies have therapeutic response. described surgical outcomes in small cohorts of older patients In the present study, basal GH levels alone had a remarkable with acromegaly (13, 24–26), all reporting surgical safety and predictive power for SRL response with an AUC of [0.72 IC: biochemical efficacy in line with that reported for younger (0.57–0.86)] increasing up to 0.82 CI (0.71–0.94) when some patients. However, there are no studies comparing first-line variables that are usually available when acromegaly is diagnosed surgery vs. SRL specifically in elderly acromegaly patients in are added. The predictive model allowed the formulation of a terms of efficacy, costs, and impact on comorbidities and on score for clinical application in which GH and gender were the quality of life, a very important endpoint especially in this subset most powerful predictors for SRL response, adjusted by IGF-1 of patients. Thus, to increase the evidence regarding first-line level, tumor size, and BMI. treatment outcomes in aged patients, it is important to perform Some other studies have also reported the response to SRL in such study in the near future, especially considering the older patients. The post-hoc analysis of the PRYMARYS clinical longevity of the general population. trial (34) examined 90 treatment-naïve acromegaly patients with Our results suggest that GH at diagnosis is a strong predictor macroadenomas, the potential predictive factor to SRL response, for SRL response in aged people suffering from acromegaly, and found that older age, female sex, and lower IGF-1 levels at particularly in women. As far as we know, this is one of the few baseline were associated with an increased probability of studies which analyze predictive factors implicated in the achieving long-term hormonal control. As in our study, female therapeutic response in older acromegaly patients. In our sex and IGF-1 were good predictors. work, neither the age (using an arbitrary cutoff of 75 years Coopmans et al. (35) analyzed 622 acromegaly patients from old) changed acromegaly clinical presentation in terms of tumor two cohorts, 194 from Rotterdam and 428 from the Liège size or GH/IGF-1 circulating concentrations, nor was the overall Acromegaly Survey Cohort treated initially with maximum response to treatment different at very advanced ages. Of note, doses of SRL. The baseline IGF-1 concentration was the best first-line medical therapy, achieved as good control as first-line predictor of good biochemical response to SRL, followed by surgery, thus reinforced the efficacy and safety of medical BMI. Younger age at diagnosis appears to be the most important treatment in this particular group of patients. In this regard, determinant of not responding, and a significant inverse we found that patients with GH lower than 6 ng/ml at diagnosis association was seen between GH concentration at diagnosis had a milder acromegaly phenotype, not just biochemically but and absolute IGF-1 reduction after treatment. Recently, Nista also regarding tumor burden (smaller and less extrasellar et al. (36) retrospectively evaluated the potential predictors of tumor), and more patients achieve disease control with SRL SRL response in a cohort of 55 naïve acromegaly patients aged 57 (27). However, differently than what has been reported by others (IQR: 45–63) and found that dichotomized age, IGF1 ULN at (28, 29), basal GH levels at diagnosis were not predictive of diagnosis, and T2-hypointense MRI signal of the tumor were surgical remission in our series; our data may be in part reliable predictors of SRL response. However, this study was not explained by the fact that surgical treatment was performed by fully comparable to ours, as patients were much younger. many different neurosurgical teams as the series consisted of In our cohort of elderly patients, low basal GH and female patients from all major pituitary centers in Spain. Previous gender were the most powerful predictors of SRL response. IGF- studies showed a correlation between tumor size and mean 1, BMI, and tumor size also contributed with less etiologic GH levels (30, 31). This could suggest that a smaller tumor fraction and were adjusting factors. Regarding tumor size, in and a lower GH at the time of diagnosis identify patients at an our cases the presence of a macroadenoma showed a statistical earlier stage of the acromegaly natural history, or alternatively, trend for good response, although it should be noted that, that their tumor is per se of low grade. However, most of these despite having up to 60% of macroadenomas, the mean tumor studies found that in the group with lower GH the patients were size was only of 12.3 ± 6.4 mm (only four patients had tumors older (31–33). Likewise, in the study of Espinosa de los Monteros larger than 25 mm), and invasiveness was low in our cohort. The et al. (31), the frequency of symptoms, signs, and comorbidities knowledge of these results will allow physicians to better advise was similar between GH groups, against the concept that low their elderly patients when they ask about surgery vs. SRL GH levels represent simply an early acromegaly disease response, especially in the subgroup of patients that refuse or detection. Similar to this study (31), in our cohort male gender have contraindication for having surgery, thus helping to share was more frequently represented in the GH <6-ng/ml group in the decision-making process. contrast with previously published series (31–33). In the low GH Our study has some limitations: it is of retrospective nature and <6-ng/dl group, a higher percentage of patients achieved disease therefore prone to introduce some bias, as patients were recruited in control with SRL compared to patients with higher basal GH, different centers in which clinical practice is based on clinical thus reinforcing the concept that acromegaly is a heterogeneous judgement according to clinical guidelines and a predefined disease even at older ages and that the disease characteristics per protocol was therefore not used. GH (and IGF-1) assays had been Frontiers in Endocrinology 08 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 modified (and eventually improved) in the time period analyzed. MP-D All the authors participated in the data collection and The strength of this study lies in the number of patients included, contributed to the article and approved the submitted version. since it is among the largest studies reported with a nationwide distribution. In addition, the analyses performed used clinical data allowing comparability and usefulness for clinical practice. Conflict of interest In conclusion, the most frequent phenotype in elderly people with acromegaly includes relatively small adenomas and moderately BB, MP, AV, CA-E, AV, RC, IB, SG, FC, SW, and MP-D high IGF-1 levels. GH at diagnosis ≤6 ng/ml and female gender, but have received consulting fees or research support or participated not age per se, were associated with greater chance of response to in clinical trials supported by from Pfizer, Ipsen, Quiasma, and/ SRL, thus confirming the heterogeneous nature of the disease, or Recordati. which also applies to patients aged 65 or older. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict Data availability statement of interest. Authors agree to make data and materials supporting the results or analyses presented in their paper available upon reasonable request. Publisher’s note All claims expressed in this article are solely those of the Ethics statement authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the The studies involving human participants were reviewed and reviewers. Any product that may be evaluated in this article, or approved by ethics committees of the Vall d’Hebron University claim that may be made by its manufacturer, is not guaranteed Hospital (number: PR(AG)318/2021). The ethics committee or endorsed by the publisher. waived the requirement of written informed consent for participation. Supplementary material Author contributions The Supplementary Material for this article can be found Conceptualization, BB, MP-D; methodology, BB; writing— online at: https://www.frontiersin.org/articles/10.3389/ original draft preparation, BB. writing—review and supervision, fendo.2022.984877/full#supplementary-material References 1. Herná ndez-Ramı́rez LC. Potential markers of disease behavior in acromegaly 8. Giustina A, Barkan A, Beckers A, Biermasz N, Biller BMK, Boguszewski C, and gigantism. Expert Rev Endocrinol Metab (2020) 15:171–83. doi: 10.1080/ et al. A consensus on the diagnosis and treatment of acromegaly comorbidities: An 17446651.2020.1749048 update. J Clin Endocrinol Metab (2019) 105(4):dgz096. doi: 10.1210/clinem/dgz096 2. Puig-Domingo M, Marazuela M. Precision medicine in the treatment of 9. Giustina A, Barkhoudarian G, Beckers A, Ben-Shlomo A, Biermasz N, Biller acromegaly. Minerva Endocrinol (2019) 44:169–75. doi: 10.23736/S0391- B, et al. Multidisciplinary management of acromegaly: A consensus. Rev Endocr 1977.18.02937-1 Metab Disord (2020) 21:667–78. doi: 10.1007/s11154-020-09588-z 3. Kasuki L, Wildemberg LE, Gadelha MR. MANAGEMENT OF 10. Ceccato F, Barbot M, Lizzul L, Cuccarollo A, Selmin E, Merante Boschin I, ENDOCRINE DISEASE: Personalized medicine in the treatment of et al. Clinical presentation and management of acromegaly in elderly patients. acromegaly. Eur J Endocrinol (2018) 178:R89–R100. doi: 10.1530/EJE-17-1006 Hormones (Athens) (2020) 20 (1):143–50. doi: 10.1007/s42000-020-00235-5 4. Cuevas-Ramos D, Carmichael JD, Cooper O, Bonert VS, Gertych A, 11. Burton T, Le Nestour E, Neary M, Ludlam WH. Incidence and prevalence of Mamelak AN, et al. A structural and functional acromegaly classification. J Clin acromegaly in a large US health plan database. Pituitary (2016) 19:262–7. Endocrinol Metab (2015) 100:122–31. doi: 10.1210/jc.2014-2468 doi: 10.1007/s11102-015-0701-2 5. Tanimoto K, Hizuka N, Fukuda I, Takano K, Hanafusa T. The influence of 12. Colao A, Pivonello R, Spinelli L, Galderisi M, Auriemma RS, Galdiero M, age on the GH-IGF1 axis in patients with acromegaly. Eur J Endocrinol (2008) et al. A retrospective analysis on biochemical parameters, cardiovascular risk and 159:375–9. doi: 10.1530/EJE-08-0243 cardiomyopathy in elderly acromegalic patients. J Endocrinol Invest (2007) 30:497– 6. Katznelson L, Laws ER, Melmed S, Molitch ME, Murad MH, Utz A, et al. 506. doi: 10.1007/BF03346334 Acromegaly: An endocrine society clinical practice guideline. J Clin Endocrinol 13. Arita K, Hirano H, Yunoue S, Fujio S, Tominaga A, Sakoguchi T, et al. Treatment Metab (2014) 99:3933–51. doi: 10.1210/jc.2014-2700 of elderly acromegalics. Endocr J (2008) 55:895–903. doi: 10.1507/endocrj.k08e-117 7. Melmed S, Bronstein MD, Chanson P, Klibanski A, Casanueva FF, Wass 14. Ambrosio MR, Gagliardi I, Chiloiro S, Ferreira AG, Bondanelli M, JAH, et al. A consensus statement on acromegaly therapeutic outcomes. Nat Rev Giampietro A, et al. Acromegaly in the elderly patients. Endocrine (2020) 68:16– Endocrinol (2018) 14:552. doi: 10.1038/s41574-018-0058-5 31. doi: 10.1007/s12020-020-02206-7 Frontiers in Endocrinology 09 frontiersin.org Biagetti et al. 10.3389/fendo.2022.984877 15. Gagliardi I, Chiloiro S, Vallillo M, Bondanelli M, Volpato S, Giampietro A, 28. Coopmans EC, Postma MR, Wolters TLC, van Meyel SWF, Netea-Maier R, et al. Multidimensional geriatric evaluation in acromegaly: A comparative cross- van Beek AP, et al. Predictors for remission after transsphenoidal surgery in sectional study. BMC Geriatr (2021) 21:598. doi: 10.1186/s12877-021-02549-4 acromegaly: A Dutch multicenter study. J Clin Endocrinol Metab (2021) 106:1783– 92. doi: 10.1210/clinem/dgab069 16. Maione L, Chanson P. National acromegaly registries. Best Pract Res Clin Endocrinol Metab (2019) 33:101264. doi: 10.1016/j.beem.2019.02.001 29. Anthony JR, Alwahab UA, Kanakiya NK, Pontell DM, Veledar E, Oyesiku NM, et al. SIGNIFICANT ELEVATION OF GROWTH HORMONE LEVEL 17. Petrossians P, Daly AF, Natchev E, Maione L, Blijdorp K, Sahnoun- IMPACTS SURGICAL OUTCOMES IN ACROMEGALY. Endocr Pract (2015) Fathallah M, et al. Acromegaly at diagnosis in 3173 patients from the liège 21:1001–9. doi: 10.4158/EP14587.OR acromegaly survey (LAS) database. Endocr Relat Cancer (2017) 24:505–18. doi: 10.1530/ERC-17-0253 30. Reid TJ, Post KD, Bruce JN, Kanibir MN, Reyes-Vidal CM, Freda PU. Features at diagnosis of 324 patients with acromegaly did not change from 1981 to 18. Dal J, Feldt-Rasmussen U, Andersen M, Kristensen LØ, Laurberg P, 2006; acromegaly remains under-recognized and under-diagnosed. Clin Endocrinol Pedersen L, et al. Acromegaly incidence, prevalence, complications and long- (Oxf) (2010) 72:203–8. doi: 10.1111/j.1365-2265.2009.03626.x term prognosis: A nationwide cohort study. Eur J Endocrinol (2016) 175:181–90. doi: 10.1530/EJE-16-0117 31. Espinosa de Los Monteros AL, Sosa-Eroza E, Gonzalez B, Mendoza V, Mercado M. Prevalence, clinical and biochemical spectrum, and treatment 19. Portocarrero-Ortiz LA, Vergara-Lopez A, Vidrio-Velazquez M, Uribe-Diaz outcome of acromegaly with normal basal GH at diagnosis. J Clin Endocrinol AM, Garcı́a-Dominguez A, Reza-Albarrá n AA, et al. The Mexican acromegaly Metab (2018) 103:3919–24. doi: 10.1210/jc.2018-01113 registry: Clinical and biochemical characteristics at diagnosis and therapeutic 32. Butz LB, Sullivan SE, Chandler WF, Barkan AL. “Micromegaly”: An update outcomes. J Clin Endocrinol Metab (2016) 101:3997–4004. doi: 10.1210/jc.2016- on the prevalence of acromegaly with apparently normal GH secretion in the 1937 modern era. Pituitary (2016) 19:547–51. doi: 10.1007/s11102-016-0735-0 20. Mestron A, Webb SM, Astorga R, Benito P, Catala M, Gaztambide S, et al. 33. Dimaraki EV, Jaffe CA, DeMott-Friberg R, Chandler WF, Barkan AL. Epidemiology, clinical characteristics, outcome, morbidity and mortality in Acromegaly with apparently normal GH secretion: implications for diagnosis and acromegaly based on the Spanish acromegaly registry (Registro espanol de follow-up. J Clin Endocrinol Metab (2002) 87:3537–42. doi: 10.1210/jcem.87.8.8658 acromegalia, REA). Eur J Endocrinol (2004) 151:439–46. doi: 10.1530/eje.0.1510439 34. Petersenn S, Houchard A, Sert C, Caron PJthe PRIMARYS Study Group. 21. Giraldi EA, Veledar E, Oyesiku NM, Ioachimescu AG. Incidentally detected Predictive factors for responses to primary medical treatment with lanreotide acromegaly: single-center study of surgically treated patients over 22 years. J autogel 120 mg in acromegaly: Post hoc analyses from the PRIMARYS study. Investig Med (2021) 69:351–7. doi: 10.1136/jim-2020-001605 Pituitary (2020) 23:171–81. doi: 10.1007/s11102-019-01020-3 22. Sesmilo G, Gaztambide S, Venegas E, Picó A, Del Pozo C, Blanco C, et al. 35. Coopmans EC, Korevaar TIM, van Meyel SWF, Daly AF, Chanson P, Brue Changes in acromegaly treatment over four decades in Spain: Analysis of the T, et al. Multivariable prediction model for biochemical response to first-generation Spanish acromegaly registry (REA). Pituitary (2013) 16:115–21. doi: 10.1007/ somatostatin receptor ligands in acromegaly. J Clin Endocrinol Metab (2020) 105: s11102-012-0384-x dgaa387. doi: 10.1210/clinem/dgaa387 23. Freda PU, Katznelson L, van der Lely AJ, Reyes CM, Zhao S, Rabinowitz D. 36. Nista F, Corica G, Castelletti L, Khorrami K, Campana C, Cocchiara F, et al. Long-acting somatostatin analog therapy of acromegaly: a meta-analysis. J Clin Clinical and radiological predictors of biochemical response to first-line treatment Endocrinol Metab (2005) 90:4465–73. doi: 10.1210/jc.2005-0260 with somatostatin receptor ligands in acromegaly: A real-life perspective. Front 24. Sasagawa Y, Hayashi Y, Tachibana O, Nakagawa A, Oishi M, Takamura T, Endocrinol (2021) 12:677919. doi: 10.3389/fendo.2021.677919 et al. Transsphenoidal surgery for elderly patients with acromegaly and its outcomes: Comparison with younger patients. World Neurosurg (2018) 118: COPYRIGHT e229–34. doi: 10.1016/j.wneu.2018.06.165 © 2022 Biagetti, Iglesias, Villar-Taibo, Moure, Paja, Araujo-Castro, Ares, ́ Alvarez-Escola, ́ Vicente, Alvarez Guivernau, Novoa-Testa, Guerrero Perez, 25. Puchner MJ, Knappe UJ, Lüdecke DK. Pituitary surgery in elderly patients with acromegaly. Neurosurgery (1995) 36:677–83; discussion 683-684. Cam ́ ara, Lecumberri, Garcıá Gom ́ ez, Bernabeu ́ , Manjon ́ , Gaztambide, doi: 10.1227/00006123-199504000-00006 Cordido, Webb, Menen ́ dez-Torre, Dıez,́ Simó and Puig-Domingo. This is 26. Minniti G, Jaffrain-Rea ML, Esposito V, Santoro A, Moroni C, Lenzi J, et al. an open-access article distributed under the terms of the Creative Surgical treatment and clinical outcome of GH-secreting adenomas in elderly Commons Attribution License (CC BY). The use, distribution or patients. Acta Neurochir (Wien) (2001) 143:1205–11. doi: 10.1007/s007010100015 reproduction in other forums is permitted, provided the original author 27. van der Lely AJ, Harris AG, Lamberts SW. The sensitivity of growth (s) and the copyright owner(s) are credited and that the original hormone secretion to medical treatment in acromegalic patients: influence of age publication in this journal is cited, in accordance with accepted and sex. Clin Endocrinol (Oxf) (1992) 37:181–5. doi: 10.1111/j.1365- academic practice. No use, distribution or reproduction is permitted 2265.1992.tb02304.x which does not comply with these terms. Frontiers in Endocrinology 10 frontiersin.org