TYPE Original Research PUBLISHED 08 August 2022 DOI 10.3389/fendo.2022.934347 Impaired quality of life, but not OPEN ACCESS cognition, is linked to a history EDITED BY Agnieszka Pazderska, St. James’s Hospital, Ireland of chronic hypercortisolism in REVIEWED BY Rosemary Dineen, patients with Cushing’s disease Mater Misericordiae University Hospital, Ireland Zhanna Belaya, in remission Endocrinology Research Center, Russia Pablo Abellán Galiana, Emilie Pupier 1, Alicia Santos 2,3, Nicole Etchamendy 4, Hospital General Universitari de Aurélie Lavielle 1, Amandine Ferriere 1, Aline Marighetto 4, Castelló(HGUCS), Spain Eugenia Resmini 2,3, Daniela Cota 4, Susan *CORRESPONDENCE M. Webb 2,3 and Antoine Tabarin 1,4* Antoine Tabarin antoine.tabarin@chu-bordeaux.fr 1 Department of Endocrinology, Diabetes and Nutrition, CHU of Bordeaux and University of Bordeaux, Pessac, France, 2 Endocrinology Department, Centro de Investigación Biomédica en Red SPECIALTY SECTION de Enfermedades Raras (CIBERER, Unidad 747) Instituto de Salud Carlos III (ISCIII), Barcelona, This article was submitted to Spain, 3 Department Medicine, Research Center for Pituitary Diseases, Hospital Sant Pau, Institut Adrenal Endocrinology, d'Investigació Biomèdica (IIB)-Sant Pau, Universitat Autònoma de Barcelona (UAB), Barcelona, a section of the journal Spain, 4 Neurocentre Magendie, University of Bordeaux, Institut National de la Santé et de la Frontiers in Endocrinology Recherche Médicale (INSERM), Bordeaux, France RECEIVED 02 May 2022 ACCEPTED 15 July 2022 PUBLISHED 08 August 2022 CITATION Context: Impaired cognition and altered quality of life (QoL) may persist despite Pupier E, Santos A, Etchamendy N, long-term remission of Cushing’s disease (CD). Persistent comorbidities and Lavielle A, Ferriere A, Marighetto A, Resmini E, Cota D, Webb SM and treatment modalities may account for cognitive impairments. Therefore, the role Tabarin A (2022) Impaired quality of of hypercortisolism per se on cognitive sequelae remains debatable. life, but not cognition, is linked to a history of chronic hypercortisolism in patients with Cushing’s disease in Objective: To investigate whether memory and QoL are impaired after long- remission. term remission of CD in patients with no confounding comorbidity. Front. Endocrinol. 13:934347. doi: 10.3389/fendo.2022.934347 Design and Setting: Cross-sectional case-control study in two tertiary referral COPYRIGHT centers © 2022 Pupier, Santos, Etchamendy, Lavielle, Ferriere, Marighetto, Resmini, Cota, Webb and Tabarin. This is an Patients: 25 patients (44.5 ± 2.4 years) in remission from CD for 102.7 ± 19.3 Mo open-access article distributed under and 25 well-matched controls, without comorbidity or treatment liable to the terms of the Creative Commons Attribution License (CC BY). The use, impair cognition. distribution or reproduction in other forums is permitted, provided the Main Outcome Measure(s): Hippocampus- and prefrontal cortex-dependent original author(s) and the copyright owner(s) are credited and that the memory, including memory flexibility and working memory, were investigated original publication in this journal is using multiple tests including sensitive locally-developed computerized tasks. cited, in accordance with accepted Depression and anxiety were evaluated with the MADRS and HADS academic practice. No use, distribution or reproduction is questionnaires. QoL was evaluated with the SF-36 and CushingQoL permitted which does not comply with questionnaires. The intensity of CD was assessed using mean urinary free these terms. cortisol and a score for clinical symptoms. Results: CD patients displayed similar performance to controls in all cognitive tests. In contrast, despite the absence of depression and a minimal residual clinical Cushing score, patients had worse QoL. Most of the SF36 subscales and Frontiers in Endocrinology 01 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 the CushingQoL score were negatively associated only with the duration of exposure to hypercortisolism (p≤ 0.01 to 0.001). Conclusions: Persistent comorbidities can be a primary cause of long-lasting cognitive impairment and should be actively treated. Persistently altered QoL may reflect irreversible effects of hypercortisolism, highlighting the need to reduce its duration. Clinical Trial Registration number: https://clinicaltrials.gov, identifier NCT02603653 KEYWORDS Cushing’s disease, hypercortisolism, cognition, memory, quality of life Introduction hypercortisolism after its cessation on cognition remain debatable. In addition, identifying the association of Chronic hypercortisolism significantly impacts brain confounding factors with persistent cognitive deficits is function and morphology (1–5). Cognition is frequently clinically relevant, because they may lead, along with somatic negatively affected during Cushing’s disease (CD) including sequelae, to a lasting deterioration in quality of life (QoL) (2–4, deficits in memory, verbal and visual learning (4, 6, 7). 6, 7, 24, 25). To address this, we investigated memory function Imaging has revealed cerebral atrophy and functional changes with multiple tests, including sensitive computerized tests for in areas rich in glucocorticoid receptors, involved in processing declarative and working memory, and QoL in patients with of cognitive functions, such as the hippocampus, amygdala, and long-term remission of CD, with no confounding comorbidity. prefrontal cortex (1, 8). Remission of CD is usually associated with cognitive improvements (4, 6, 7, 9), but controversies persist as to the degree of recovery (4, 6, 7). Numerous factors Patients and methods may account for the heterogeneity between the results of studies. These include variations in recruitment, duration of Patients hypercortisolism and time elapsed between remission of CS and cognitive evaluation, assuming that cognitive recovery Clinical charts of 192 patients with cured CD and followed may be delayed (2, 10, 11), discrepancies between cross- for > 1 year were reviewed in two expert centers (Endocrinology sectional and longitudinal studies, and differences in departments of Haut Leveque Hospital, Bordeaux, France and neuropsychological tests used (6, 7, 11, 12). Besides these Hospital Sant Pau, Barcelona, Spain). CD was confirmed on caveats, variable memory recovery between individuals histopathological analysis, remission of CD following questions the specific role of hypercortisolism and suggests transsphenoidal surgery, and the results of bilateral inferior that different causes, directly or indirectly linked to petrosal sinus sampling (BIPSS). Patients were selected using hypercortisolism, may be involved in persistent alterations (12, two main criteria: 1) being in definitive remission following 13). Indeed, patients may have confounding factors impairing surgery for at least one year and 2) the absence of confounding cognition and, more specifically, memory. These include factors or comorbidity that may alter cognitive functions. advanced age (14), obesity, poorly-controlled diabetes (15, 16), Remission was ascertained during the 3 months preceding cerebral vascular disease and poorly controlled hypertension cognitive evaluation and was defined by persistent adrenal (17), psychopathology and depression including psychotropic insufficiency or normal 24h urinary free cortisol excretion treatments (18), excessive hydrocortisone replacement, (UFC) associated with serum cortisol < 50 nmol/L following 1 imperfect hormonal replacement for hypothyroidism and mg overnight dexamethasone suppression test. Patients with untreated GH deficiency (19–21). Among the treatment adrenal insufficiency had to receive hydrocortisone modalities, pituitary radiotherapy may be associated to supplementation at doses ≤ 15 mg/m2. Conditions known to cognitive deficits (22, 23). A number of these putative interfere with cognitive function (considered exclusion criteria) interfering factors were not controlled or specified in prior included: age > 60 years, remission of CD induced by drugs, publications. Therefore, the specific long-term consequences of previous pituitary radiotherapy, BMI > 30 kg/m², drugs- and Frontiers in Endocrinology 02 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 alcohol-abuse (past and present), untreated GH-deficiency or declarative memory; and organization of information in hypothyroidism, uncontrolled diabetes (HbA1c > 7.5% or FBG > memory and sensibility to interference (30–32) (see 1.40 g/L) or diabetes-induced organ damage, past-history of Supplemental Data). neurological/vascular disease including uncontrolled hypertension, intake of psychotropic drugs including sleeping tablets and intake of exogenous steroids. GH secretion was Evaluation of depression and anxiety assessed using dynamic testing and was considered normal in presence of a GH peak > 3 ng/mL during insulin tolerance test Depression and anxiety were assessed using the Depression and glucagon stimulation test or a GH peak > 8 ng/mL during Rating Scale (MADRS) questionnaire and the Hospital Anxiety the GHRH -arginine stimulation test. and Depression Scale (HADS) (33, 34). Control group Evaluation of QoL Each patient with CD recruited a control subject matched for The generic Short Form 36 (SF-36) questionnaire (35) was age, gender, living area, socioeconomic status and educational used in both patients and controls. The specific CushingQoL level. Five educational levels were defined: 1: no school, 2: questionnaire was also used in patients (36). primary school, 3: 5-9 years at school, short secondary school, 4: more than 9 years at school, long secondary school, 5: more than 12 years, university level. Exclusion criteria for controls Evaluation of Cushing’s disease were identical to those of CD patients; thyroid function was evaluated, but GH secretion was not. We used an arbitrarily defined score for symptoms of Informed consent was obtained from all participants and the hypercortisolism to evaluate clinical intensity of CD. The score study was approved by the ethical committees of both centers. The was calculated following clinical examination at the time of study was registered under Clinical Trials (ClinicalTrials.gov cognitive evaluation and retrospectively calculated from medical Identifier: NCT02603653). reports at diagnosis. A score of 1 was recorded for the presence of any symptom like hirsutism, menstrual irregularities, buffalo neck, facial plethora and central obesity, diabetes, dyslipidemia, Methods hypertension and cognitive or psychopathological complaints. A score of 2 was recorded for the presence of more specific All evaluations for the patients and the controls were symptoms such as large purple striae, proximal muscle performed at the same time during one day. weakness, spontaneous ecchymosis and glucocorticoid-induced osteoporosis. The maximum score was 20. The biological intensity of CD at diagnosis was calculated using the mean Cognitive evaluation results of 2-3measurements of 24h UFC expressed relative to the upper normal range of the assay (x ULN). Cognitive and psychological evaluation was performed by a The apparent duration of hypercortisolism was calculated dedicated psychologist to explore memory (hippocampal- and retrospectively from the time between onset of symptoms frontal- dependent forms), fluency, mood (including depression according to the patient interview, and the date of remission and anxiety); and QoL. The neuropsychological outcomes were of hypercortisolism. explored using: 1) The Rey Auditory Verbal Learning Test (RAVLT) to evaluate verbal episodic memory (26), 2) the Isaac’ Set tests assessing phonemic and semantic verbal fluency Statistical analysis abilities and speed of verbal production (27), 3) two subtests of the Wechsler Memory Scale, 4th Edition that evaluates visual A power analysis was performed: according to a hypothesis and visuo-spatial working memory: the spatial addition task and that the percentage of correct answers to the Bordeaux maze test the symbol span task (28), 4) the Rey-Osterrieth Complex Figure would be 80% in controls, with a standard deviation of 13% and (ROCF) test dedicated to the evaluation visuospatial 60% with a standard deviation of 23% in CD patients, 25 patients constructional ability and visual memory (29). Only the per group were necessary to have a power of 80% associated with delayed recall was performed in this test. a type 1 risk of 5%. Cognitive performance parameters recorded In addition, patients underwent two locally developed in each task and scores recorded in the different scales assessing computerized maze memory tests assessing cardinal features of QoL, anxiety and depression were analyzed with ANOVAs to declarative memory and working memory such as flexibility of assess potential effect of group factor. Comparison of data Frontiers in Endocrinology 03 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 between patients with recovery of the HPA axis and those with dose divided in 2-3 daily intakes in 12 patients (e.g. 10 mg/day in 3, persistent corticotropic insufficiency were analyzed using the t- 11 to 20 mg/day in 5 and 21 to 30 mg/day in 4. test or Mann-Whitney’s test according to the distribution of Thirteen patients had a BMI ≥ 25 and < 30 kg/m2. Five were values. All results are expressed as mean ± SEM. Multiple supplemented with levothyroxine and 2 with recombinant GH. correlation analysis were performed with Spearman test after Blood pressure was controlled pharmacologically in 5 hypertensive Log transformation of data using GraphPad software (San patients (≤135/85 mmHg); only one patient was treated with Diego, CA). We also compared Z-scores (computed using log antidiabetic hypoglycemic oral drugs (HbA1C = 6.8%). data for each participant taking into account the mean and SD of Mean UFC at diagnosis was 4.4 ± 0.6 x ULN (range 1.5 – each measure) of SF36 items in controls versus patients. A 12.4). An improvement in the clinical score was observed between negative z-score indicates that patients has lower quality of life diagnosis and the time of evaluation (12.6 ± 0.7 vs 2.2 ± 0.5, than the corresponding normative population. The Bonferroni respectively; p < 0.0001). Clinical scores at diagnosis and at the method was used to reduce the likelihood of type I error. Since time of evaluation were similar between patients with and without we performed 4 different evaluations (cognitive, depression, adrenal insufficiency (13.5 ± 1.0 and 11.7 ± 0.9; p=0.2 and 2.3 ± 0.8 anxiety, and quality of life), a p value ≤0.01 was considered vs 2.2 ± 0.5, p = 0.91, respectively). as significant. Twenty-five subjects (19 women and 6 men, aged 44.3 ± 2.3 y), were recruited as controls. CD patients and controls had similar characteristics (Table 1). Apart from one individual receiving Results adequate levothyroxine supplementation for hypothyroidism, no control subject had a history of endocrine disease.no control subject Characteristics of patients and controls had a history of endocrine disease. Clinical files of 192 patients with CD were analyzed. Twenty- five (19 women and 6 men, aged 44.5 ± 2.4 y) in long-term CD patients in long-term remission have remission of CD (102.7 ± 19.3 months; range =17-364) satisfied the preserved cognitive performance inclusion criteria and accepted to participate (Table 1 and Figure 1). Ten were recruited in Barcelona and 15 in Bordeaux. Twenty-two CD patients in long-term remission displayed similar were cured following a single pituitary surgery. Among these, 13 performance to controls in all cognitive tests. Regarding patients had a normal biological evaluation of the HPA axis and 9 declarative memory, total words recalled in the RAVLT test was had persistent corticotropic insufficiency. Three underwent bilateral not different between control and CD groups [group effect: F adrenalectomy following 1 (n=1) or 2 (n=2) unsuccessful pituitary (1,48)=1.81; p=0.18] and CD patients in remission performed as surgical attempts. Hydrocortisone was given at 11.3 ± 1.9mg/m² controls in all RAVLT subtests [i.e. delayed recall: F (1,48)=1.7; TABLE 1 Characteristics of participants in the Memocush study. Patients (n=25) Controls (n=25) Age (years) 44.5 +/- 2.4 44.3 +/- 2.3 Women (%) 76 76 Level of education (from 1 to 5) Level 4 (n=8) Level 4 (n=7) & Level 5 (n=17) & Level 5 (n=18) BMI (kg/m2) 24.2 +/- 0.6 23.1 +/- 0.6 Treated hypertension (N) 5 0 Systolic blood pressure (mm Hg) 123.2 +/- 3.2 116.6 +/- 2.6 Diastolic blood pressure (mm Hg) 74.4 +/- 2.0 71.3 +/- 2.3 Diabetes mellitus (N) 1 0 Fasting glycemia (mmol/l) 4.9 +/- 0.1 5.1 +/- 0.1 HbA1c (%) 5.4 +/- 0.1 5.4 +/- 0.1 Hydrocortisone treatment (Number of patients) 12 0 Hydrocortisone total dose (mg/m² per day) 11.3 +/- 1.9 NA Treated Hypothyroidism (Number of patients) 5 1 Treated GH deficiency (Number of patients) 2 0 Clinical score at diagnosis (/20) 12.6 +/- 0.7 NA Clinical score at evaluation (/20) 2.3 +/- 0.5 NA No significant difference was found in the non-specific parameters between patients with a history of Cushing’s disease and their matched controls. Frontiers in Endocrinology 04 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 FIGURE 1 Flow chart of the recruitment of patients with Cushing’s disease (CD) in the two investigating centers. p=0.19, for all the subtests, mean and p value in Table 2]. In the group: F (5,240)=1.2; p=0.19). When submitted to flexibility Bordeaux Maze declarative memory test, patients learned probes, CD patients accurately expressed their memories of accurately the initial 6 pairs, needing similar number of training previously acquired information similarly to controls [group sessions as controls [group effect: F (1,48)=1.7; p=0.19]. When effect: F (1,48)=0.7; p=0.39]. 6/25 CD patients displayed considering the evolution of performance during the last 6 trials performance in the flexibility probes below or equal to 62.5% of preceding achievement of criterion performance, CD patients correct responses, similar to that observed in the controls. progressively reached a mean percent of correct choices (96%) Regarding working memory, symbol span score was not close to that of controls (98%), with no difference between both different between CD patients and controls [group effect: F (1,48) groups (group effect: F (1,48)=1.8; p=0.31; interaction trials X =0.2; p=0.63]. In the spatial addition subtest, CD patients in TABLE 2 Mean responses during various cognitive tests. Controls CD p value List A-1 7.1 ± 0.5 (1/25) 6.4 ± 0.5 (5/25) p=0.24 List A-2 10.6 ± 0.6 (1/25) 9.7 ± 0.5 (1/25) p=0.26 List A-3 12.0 ± 0.6 (1/25) 11.0 ± 0.4 (1/25) p=0.16 List A-4 12.7 ± 0.4 (1/25) 12.2 ± 0.4 (1/25) p=0.4 List A-5 13.2 ± 0.3 (1/25) 12.7 ± 0.4 (1/25) p=0.31 List A-6-Retention 12.0 ± 0.5 (1/25) 10.7 ± 0.6 (1/25) p=0.09 Delayed A 12.2 ± 0.5 (1/25) 11.2 ± 0.5 (1/25) p=0.19 P 26.0 ± 1.3 (0/25) 26.6 ± 1.3 (0/25) p=0.76 R 22.5 ± 1.3 (2/25) 20.5 ± 1.3 (1/25) p=0.28 V 16.5 ± 0.9 (1/25) 17.2 ± 1.0 (0/25) p=0.59 Animal 33.5 ± 1.5 (2/25) 33.6 ± 1.6 (0/25) p=0.94 Fruit 17.0 ± 1.0 (9/25) 16.5 ± 1.0 (7/25) p=0.74 Furniture 21.7 ± 1.0 (0/25) 21.4 ± 0.7 (0/25) p=0.27 Spatial Addition 14.1 ± 0.8 (4/25) 13.7 ± 0.7 (4/25) p=0.69 Symbol 25.2 ± 1.3 (3/25) 26.2 ± 1.6 (4/25) p=0.63 ROCF Copy 55.6 ± 3.6 (0/25) 54.6 ± 3.6 (0/25) p=0.87 ROCF delayed 33.1 ± 2.6 (1/25) 31.4 ± 2.6 (1/25) p=0.1 Correct responses in the successive free recall subtests of the RAVLT (26); words produced for each letter or category subtest of the Isaac Set Test (27), performance in the spatial addition test, the symbol test and the ROCF test (29). For each parameter, the number of deficient participants according to the published normative value is specified and the p value indicates the effect of group. CD: patients with a history of Cushing’s disease. Frontiers in Endocrinology 05 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 remission performed similarly to controls [group effect: F (1,48) F (1,48)=10.2 p<0.005; vitality: F (1,48)=6.3; p<0.01; social =0.2; p=0.69]. In the Bordeaux Working memory test, the total functioning: F (1,48)=5.1 p<0.05; role-emotional: F (1,48)=3.5; mean correct performance of the CD group was above chance level p=0.06; bodily pain: F (1,48)=1.7; p=0.19; mental health: F (1,48) and not different from controls [group effect: F (1,48)=1.9; p=0.16]. =3.04; p=0.08] (Table 3). Z scores of SF36 (especially physical Regarding verbal fluency, CD patients in long-term functioning, general Health and vitality) were also significantly remission performed similarly to controls both in the semantic lower in CD patients than in controls (Supplementary Table 1). subtest [group effect: F (1,48)=0.2; p=0.67] and in the phonetic CD patients scored worse than controls in the depression subtest [group effect: F (1,48)=0.03; p=0.85]. rating scale [MADRS: F (1,48)=3.8; p = 0.01]. 10/25 CD patients The ROCF scores were all found to be within acceptable displayed mild (8/25) or moderate (2/25) depression as opposed limits. Performance on the ROCF copy or the ROCF recall test to 12% 3/25 of the controls (p=0.04). These mild alterations were were not different between both groups [copy: F (1,48)=0.03, not apparent in the depression subscale of the HADS p=0.87; recall: F (1,48)=0.09, p=0.76] questionnaire [F (1,48)=0.51; p=0.47]. Similarly, no difference According to published normative values for RAVLT-, was found between patients and controls in the anxiety subscale Isaacs-, spatial addition- and symbol tests, the proportion of of the HADS questionnaire [F (1,48)=2.46; p=0.12] (Table 4). patients with abnormal responses was similar in CD and The CushingQoL global score was 62.6 ± 3.9. control groups. There was no significant difference in the results of tests Patients with normal HPA axis function performed better between patients requiring hydrocortisone supplementation and than patients requiring hydrocortisone supplementation only patients with a normal HPA axis. Only a nominal association was during the Isaac lexical test (24.0 ± 1.2 vs 18.6 ± 1.1; p < 0.005). found for the general health perception of the SF-36 questionnaire (47.3 ± 7.8 vs 66.1 ± 4.9 for patients taking hydrocortisone and patients with normal HPA axis function respectively; p = 0.05). CD patients in long-term remission have substantially impaired QoL Worse QoL is correlated to the duration CD patients scored worse than controls in most subscales of of exposure to cortisol excess the SF-36 questionnaire taken independently with differences reaching statistical significance for physical functioning, role- Biological intensity of hypercortisolism was not correlated to physical, general health and vitally:[physical functioning: F (1,48) the clinical score at diagnosis, nor to the clinical score at the time =15.6; p<0.005; role-physical: F (1,48)=9.2; p<0.005; general health: of evaluation (r=0.39; p=0.10 and r=0.32; p=0.19, respectively). TABLE 3 Mean score in each subscale of the SF-36 for participants. SF36-PF SF36-RP SF36-BP SF36-GH SF36-VIT SF36-SF SF36-RE SF36-MH Controls 94.8 ± 1.6 87.7 ± 5.1 86.1 ± 3.8 74.8 ± 2.7 70.0 ± 3.7 92.4 ± 3.8 90.7 ± 4.4 81.2 ± 2.9 2/25 2/25 1/25 1/25 3/25 3/25 1/25 1/25 CD 78.9 ± 3.7 58.9 ± 7.1 78.5 ± 4.4 57.0 ± 4.8 54.2 ± 5.1 77.8 ± 5.5 74.9 ± 7.2 72.7 ± 3.8 10/25 8/25 1/25 7/25 12/25 6/25 5/25 3/25 p value p=0.0003 p=0.0039 p=0.19 p=0.0024 p=0.015 p=0.03 p=0.068 p=0.087 CD, patients with a history of Cushing’s disease. For each parameter; the number of deficient participants according to the published normative value is specified and the p value indicates the effect of group. PF, physical functioning; RP, Role Physical; BP, Bodily Pain; GH, General Health; VIT, Vitality; SF, Social Function; RE, Role Emotional; MH, Mental Health. TABLE 4 Mean score in each subscale of the Hospital Anxiety and Depression Scale (HADS) and the Depression Rating Scale (MADRS) for each group of participants. HADS Depression HADS Anxiety MADRS Controls 3.1 ± 0.7 6.6 ± 0.7 5.0 ± 1.4 Doubtful cases : 1/25 Doubtful cases : 1/25 Mild depression : 2/25 Certain cases : 1/25 Certain cases : 4/25 Moderate depression : 1/25 CD 3.8 ± 0.6 7.4 ± 0.7 9.8 ± 1.1 Doubtful cases : 3/25 Doubtful cases : 3/25 Mild depression : 8/35 Certain cases : 0/25 Certain cases : 7/25 Moderate depression : 2/25 p value p=0.47 p=0.12 p=0.04 CD, patients with a history of Cushing’s disease. For each test, the number of participants with abnormal score is specified according to normative values and the p value indicates the effect of group. Frontiers in Endocrinology 06 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 None of the QoL, cognition and anxiety/depression evaluations dimension of the SF-36 questionnaire was similar to that of including all SF-36 subscales correlated with the biological controls. Given the debate on the consequences of brain intensity of hypercortisolism and clinical score at diagnosis (p irradiation, patients after pituitary radiotherapy were excluded = 0.16 to 0.99). In contrast, the general health, vitality, role (9, 22). Hydrocortisone replacement dosage and GH deficiency physical and physical functioning subscales of the SF-36, as well were also controlled to avoid interference between imperfect as CushingQoL were significantly and nominally negatively hormonal replacement and cognitive performance (19–21). A associated with the duration of exposure to hypercortisolism: delay between remission of CD and recovery of cognitive r=-0.54, p= 0.004; r=-0.61, p=0.001; r=-0.53, p=0.01; r=-0.49, impairments has also been suggested (2, 10). The longer p=0.02; r=-0.43, p=0.04; respectively. No parameter was duration of remission (mean = 8.6 years) before cognitive associated with the time elapsed between remission and time evaluation in our study as compared to several published of testing (p=0.19 to 0.85). studies may partly account for our findings. Interestingly, one of the largest published studies with the longest duration of remission, that also considered a number of confounding Discussion comorbidities for selecting patients, only found subtle long- term residual alterations in cognitive function and memory The main finding of our study is that, in the absence of obvious (9). The mild differences with the results of our study, in confounding comorbidity, adult patients younger than 60 years in which several similar cognitive tests were used may also be surgical remission of CD displayed no long-term memory related to some residual confounders. For instance, our patients impairment. These findings were obtained after multiple tests for were younger (44 ± 2 years vs 52 ± 1 years; P = 0.008) and had a memory evaluation, including the Bordeaux Maze relational higher educational level, two parameters associated with the memory test that has been shown to be exquisitely sensitive to outcome of most cognitive tests (9, 13, 43). In addition, 27% of mild changes in declarative memory, as occurs with ageing (30, 31, the Dutch patients underwent pituitary radiotherapy, a 37, 38). In contrast, and in the absence of obvious psychopathology treatment modality negatively associated with memory and and major physical sequelae, patients displayed impaired QoL, the executive functioning (9). intensity of which was proportional to the duration of In accordance with previous studies using generic and active hypercortisolism. disease-specific questionnaires (2, 6, 7, 24, 25, 47), CD patients Studies evaluating patients in remission of CD show variable in remission demonstrated persistent worse QoL, with a similar improvement in cognitive domains and a recent publication CushingQoL score to those previously published (48). pinpointed that the scarcity of data did not allow metanalysis of Depression and physical sequelae are major determinants of studies for memory (6). We hypothesized that cortisol-related altered QoL (48). Therefore, an intriguing finding of our study is comorbidities and treatments of CD may account for persistent the contrast between persistently altered QoL on the one hand, memory deficits and impaired QoL. This study was therefore and lack of memory impairment, evident psychopathology and carried out in an attempt to discriminate the respective impact of major residual somatic comorbidity as illustrated by the minimal a direct effect of hypercortisolism on the brain versus that of residual clinical score on the other hand. These differences other causes, directly or indirectly linked to hypercortisolism in between perception and objective findings are reminiscent of persistent memory impairment and altered QoL. the description of an increased negative illness perception in CD Thus, we strictly excluded putative confounding factors (49). Subtle psychopathological alterations and unconscious which may explain the discrepancy with more pessimistic fears related to the mental experience of hypercortisolism that studies on the recovery of cognitive functions (9–13, 39–43). cannot be detected by questionnaires cannot be excluded. Close For example, while most investigators used an age limit of 70, we links between glucocorticoids and addiction have been described decided, based on studies devoted to the decline in memory in animal models including mediation of addictive properties of performance with age (31, 37, 44), to use an age limit of 60 years. drugs by glucocorticoids (50, 51). The positive reinforcing effects Several studies included patients treated for depression. of sustained hypercortisolism may determine reward-related However, both depression and long-term intake of psychopathologies (51) and, due to their psychostimulant antidepressant drugs can induce cognitive impairments (18, properties, corticosteroids have been labelled as drugs of 45). Similarly, the use of any psychotropic drug including addiction (52). The negative emotional state and symptoms sleeping pills, described in almost 1/3 of patients with a of drug abuse withdrawal (53) are also observed after history of CD (46), was an exclusion criterion. No major remission of hypercortisolism, at least in the short-term (54). psychological disturbance was found in our patients. A trends One could hypothesize that prolonged impaired QoL may share towards increased prevalence of depression was found using the common mechanisms with those observed following withdrawal MADRS, but depression was rated as mild and patients did not from drug abuse. In this conceptual framework, it is worth differ from controls during the HADS screening questionnaire mentioning that improvements in QoL in this last situation can for depression and anxiety. Accordingly, the mental health be delayed for decades, with lower indices of well-being over Frontiers in Endocrinology 07 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 time in women with a history of psychostimulant abuse as that various persisting co-morbidities of CD may be responsible compared to alcohol and cannabis abuse (55). Improvement in for long-lasting impaired memory and should therefore be psychological care of patients with a history of CD may also actively sought and adequately treated by expert specialists benefit from neuroendocrine research in the field of addiction. (46). The contrast between lasting impairment of QoL on the Importantly, most of the SF-36 subscales and the one hand and absence of major physical, psychological and CushingQoL score were negatively associated with the cognitive sequelae on the other hand, may reflect irreversible calculated duration of exposure to hypercortisolism, but not consequences of cortisol excess. The association of negative with the duration of remission. This has also been described in a health perception and impaired QoL with the duration of retrospective survey (47) and is reminiscent of studies showing cortisol excess reinforces the importance of shortening this that the persistence of somatic and psychological cortisol-related exposure, especially following diagnosis, to reduce the long- comorbidities correlates with the delay to diagnosis (56). term burden of CD. Our study has several limitations. Being cross-sectional, it does not allow comparison with alterations observed during the active phase of CD or concluding that cognitive dysfunction in Data availability statement CD is reversible. However, a cross-sectional study avoids learning the cognitive tests that occurs with their repetition The raw data supporting the conclusions of this article will and may result in an apparent improvement of performances. In be made available by the authors, without undue reservation. addition, our study probably escapes the selection bias of longitudinal studies with high losses at follow-up, that may select patients who perform worse (6, 7). Dissociations in the Ethics statement recovery between various cognitive domains have been reported (10, 11) and we cannot exclude the persistence of other less The studies involving human participants were reviewed and frequently observed cognitive deficits, not explored within approved by CPP Sud-Ouest et Outre-Mer III and ethical our protocol. committee of hospital de la Santa creu I Sant-Pau. The The number of patients is relatively small, a problem patients/participants provided their written informed consent practically unavoidable in rare diseases, especially if followed to participate in this study. up over years, but compares well to most series related to memory performance published to date (6, 10, 12, 39, 41–43). As potent confounders are very common in patients with a Author contributions history of CD, an important proportion of patients were excluded. Although we agree that this selection excludes the AS, NE, AM, DC, and AT conceptualized the study. NE and most common situation in which patients have persistent AM developed the Bordeaux maze investigations. EP, AL, and comorbidities, it was mandatory to delineate the specific role AS performed the investigations and collected the data. NE, EP, of past hypercortisolism and that of persistent cortisol- and AT analyzed the data. AS, ER, DC, SW, and AT wrote the dependent or non-cortisol-dependent comorbidities on manuscript. AT takes responsibility for the integrity of the data cognition and QoL. However, the comparison of parameters analysis AF performed the investigations and analyzed the data. between patients with adrenal insufficiency and those with All authors contributed to the article and approved the normal HPA axis function may lack statistical power given the submitted version. small number of patients. Among the methodological merits of our study, the careful matching of patients with controls is worth mentioning, a Funding condition that likely resulted in a higher accuracy than studies involving population-based cohorts. In addition, the selection of This work was supported entirely with institutional funds controls by patients enabled a perfect match for socioeconomic and sponsored by Institut National de la Santé et de la Recherche status, which may be a determinant of the outcomes of the Mé dicale (INSERM), France. questionnaires, along with age, gender, and education level. In conclusion, our study challenges the concept of irreversible memory impairment due to a specific and direct Acknowledgments effect of hypercortisolism in patients below the age of 60. Although an increased vulnerability of the brain to cortisol We thank Prof. K Mohammedi for his assistance in the excess in aged people cannot be excluded, our results suggest statistical analyses. Frontiers in Endocrinology 08 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 Conflict of interest Supplementary material The authors declare that the research was conducted in the The Supplementary Material for this article can be found absence of any commercial or financial relationships that could online at: https://www.frontiersin.org/articles/10.3389/ be construed as a potential conflict of interest. fendo.2022.934347/full#supplementary-material Publisher’s note SUPPLEMENTARY DATA SHEET 1 Computerized maze memory tests assessing cardinal features of declarative memory and working memory. All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated SUPPLEMENTAL TABLE 1 organizations, or those of the publisher, the editors and the Z-scores of each SF36 subscale in controls and patients with Cushing’s reviewers. Any product that may be evaluated in this article, or disease in remission Z-scores of log data were computed for each participant taking into account the mean and SD of log SF36 measures. claim that may be made by its manufacturer, is not guaranteed A negative z-score indicates that the population has lower quality of life or endorsed by the publisher. than the corresponding normative population. References 1. Andela CD, van Haalen FM, Ragnarsson O, Papakokkinou E, Johannsson G, 14. DeCarli C. Mild cognitive impairment: Prevalence, prognosis, aetiology, Santos A, et al. Mechanisms in endocrinology: Cushing’s syndrome causes and treatment. Lancet Neurol (2003) 2(1):15–21. doi: 10.1016/S1474-4422(03) irreversible effects on the human brain: a systematic review of structural and 00262-X functional magnetic resonance imaging studies. Eur J Endocrinol (2015) 173(1):R1– 15. O’Brien PD, Hinder LM, Callaghan BC, Feldman EL. Neurological 14. doi: 10.1530/EJE-14-1101 consequences of obesity. Lancet Neurol (2017) 16(6):465–77. doi: 10.1016/S1474- 2. Feelders RA, Pulgar SJ, Kempel A, Pereira AM. The burden of cushing’s 4422(17)30084-4 disease: clinical and health-related quality of life aspects. Eur J Endocrinol (2012) 16. van Sloten TT, Sedaghat S, Carnethon MR, Launer LJ, Stehouwer CDA. 167(3):311–26. doi: 10.1530/EJE-11-1095 Cerebral microvascular complications of type 2 diabetes: stroke, cognitive 3. Piasecka M, Papakokkinou E, Valassi E, Santos A, Webb SM, de Vries F, et al. dysfunction, and depression. Lancet Diabetes Endocrinol (2020) 8(4):325–36. doi: Psychiatric and neurocognitive consequences of endogenous hypercortisolism. J 10.1016/S2213-8587(19)30405-X Intern Med (2020) 288(2):168–82. doi: 10.1111/joim.13056 17. Haring B, Wu C, Coker LH, Seth A, Snetselaar L, Manson JE, et al. 4. Resmini E, Santos A, Webb SM. Cortisol excess and the brain. Front Horm Hypertension, dietary sodium, and cognitive decline: Results from the women’s Res (2016) 46:74–86. doi: 10.1159/000443868 health initiative memory study. Am J Hypertens. (2016) 29(2):202–16. doi: 10.1093/ 5. Santos A, Resmini E, Gomez-Anson B, Crespo I, Granell E, Valassi E, et al. ajh/hpv081 Cardiovascular risk and white matter lesions after endocrine control of cushing’s 18. Culpepper L, Lam RW, McIntyre RS. Cognitive impairment in patients with syndrome. Eur J Endocrinol (2015) 173(6):765–75. doi: 10.1530/EJE-15-0600 depression: awareness, assessment, and management. J Clin Psychiatry (2017) 78 6. Broersen LHA, Andela CD, Dekkers OM, Pereira AM, Biermasz NR. (9):1383–94. doi: 10.4088/JCP.tk16043ah5c Improvement but no normalization of quality of life and cognitive functioning 19. Blacha AK, Rahvar AH, Flitsch J, van de Loo I, Kropp P, Harbeck B. after treatment of cushing syndrome. J Clin Endocrinol Metab (2019) 104 Impaired attention in patients with adrenal insufficiency - Impact of (11):5325–37. doi: 10.1210/jc.2019-01054 unphysiological therapy. Steroids (2021) 167:108788. doi: 10.1016/ 7. Frimodt-Moller KE, Mollegaard Jepsen JR, Feldt-Rasmussen U, Krogh J. j.steroids.2020.108788 Hippocampal volume, cognitive functions, depression, anxiety, and quality of life 20. Falleti MG, Maruff P, Burman P, Harris A. The effects of growth hormone in patients with cushing syndrome. J Clin Endocrinol Metab (2019) 104(10):4563– (GH) deficiency and GH replacement on cognitive performance in adults: a meta- 77. doi: 10.1210/jc.2019-00749 analysis of the current literature. Psychoneuroendocrinol (2006) 31(6):681–91. doi: 8. Santos A, Granell E, Gomez-Anson B, Crespo I, Pires P, Vives-Gilabert Y, 10.1016/j.psyneuen.2006.01.005 et al. Depression and anxiety scores are associated with amygdala volume in 21. Jurado-Flores M, Warda F, Mooradian A. Pathophysiology and clinical cushing’s syndrome: preliminary study. BioMed Res Int (2017) 2017:2061935. doi: features of neuropsychiatric manifestations of thyroid disease. J Endocrine Soc 10.1155/2017/2061935 (2022) 6(2):bvab194. doi: 10.1210/jendso/bvab194 9. Tiemensma J, Kokshoorn NE, Biermasz NR, Keijser BJ, Wassenaar MJ, 22. Castinetti F, Brue T, Ragnarsson O. Radiotherapy as a tool for the treatment Middelkoop HA, et al. Subtle cognitive impairments in patients with long-term of cushing’s disease. Eur J Endocrinol (2019) 180(5):D9–D18. doi: 10.1530/EJE-19- cure of cushing’s disease. J Clin Endocrinol Metab (2010) 95(6):2699–714. doi: 0092 10.1210/jc.2009-2032 23. Heald AH, Ghosh S, Bray S, Gibson C, Anderson SG, Buckler H, et al. Long- 10. Forget H, Lacroix A, Bourdeau I, Cohen H. Long-term cognitive effects of term negative impact on quality of life in patients with successfully treated glucocorticoid excess in cushing’s syndrome. Psychoneuroendocrinol (2016) 65:26– cushing’s disease. Clin Endocrinol (Oxf) (2004) 61(4):458–65. doi: 10.1111/ 33. doi: 10.1016/j.psyneuen.2015.11.020 j.1365-2265.2004.02118.x 11. Hook JN, Giordani B, Schteingart DE, Guire K, Giles J, Ryan K, et al. 24. Lindsay JR, Nansel T, Baid S, Gumowski J, Nieman LK. Long-term Patterns of cognitive change over time and relationship to age following successful impaired quality of life in cushing’s syndrome despite initial improvement treatment of cushing’s disease. J Int Neuropsychol Soc: JINS (2007) 13(1):21–9. after surgical remission. J Clin Endocrinol Metab (2006) 91(2):447–53. doi: doi: 10.1017/S1355617707070051 10.1210/jc.2005-1058 12. Starkman MN, Giordani B, Gebarski SS, Schteingart DE. Improvement in 25. Santos A, Resmini E, Martinez Momblan MA, Valassi E, Martel L, Webb learning associated with increase in hippocampal formation volume. Biol SM. Quality of life in patients with cushing’s disease. Front Endocrinol (2019) Psychiatry (2003) 53(3):233–8. doi: 10.1016/S0006-3223(02)01750-X 10:862. doi: 10.3389/fendo.2019.00862 13. Resmini E, Santos A, Gomez-Anson B, Vives Y, Pires P, Crespo I, et al. 26. Rosenberg SJ, Ryan JJ, Prifitera A. Rey auditory-verbal learning test Verbal and visual memory performance and hippocampal volumes, measured by 3- performance of patients with and without memory impairment. J Clin Psychol Tesla magnetic resonance imaging, in patients with cushing’s syndrome. J Clin (1984) 40(3):785–7. doi: 10.1002/1097-4679(198405)40:3<785::AID- Endocrinol Metab (2012) 97(2):663–71. doi: 10.1210/jc.2011-2231 JCLP2270400325>3.0.CO;2-4 Frontiers in Endocrinology 09 frontiersin.org Pupier et al. 10.3389/fendo.2022.934347 27. Gleissner U, Elger CE. The hippocampal contribution to verbal fluency in treatment for cushing’s disease and acromegaly. J Endocrinol Invest (2011) 34(7): patients with temporal lobe epilepsy. Cortex (2001) 37(1):55–63. doi: 10.1016/ e168–77. doi: 10.3275/7333 S0010-9452(08)70557-4 42. Martignoni E, Costa A, Sinforiani E, Liuzzi A, Chiodini P, Mauri M, et al. 28. Tulsky DS. A new look at the WMS-III: new research to guide clinical The brain as a target for adrenocortical steroids: cognitive implications. practice. J Clin Exp Neuropsychol (2004) 26(4):453–8. doi: 10.1080/ Psychoneuroendocrinol (1992) 17(4):343–54. doi: 10.1016/0306-4530(92)90040-E 13803390490490773 43. Mauri M, Sinforiani E, Bono G, Vignati F, Berselli ME, Attanasio R, et al. 29. Liberman J, Stewart W, Seines O, Gordon B. Rater agreement for the rey- Memory impairment in cushing’s disease. Acta Neurol Scand (1993) 87(1):52–5. osterrieth complex figure test. J Clin Psychol (1994) 50(4):615–24. doi: 10.1002/ doi: 10.1111/j.1600-0404.1993.tb04075.x 1097-4679(199407)50:4<615::AID-JCLP2270500419>3.0.CO;2-R 44. Marighetto A, Brayda-Bruno L, Etchamendy N. Studying the impact of 30. Etchamendy N, Konishi K, Pike GB, Marighetto A, Bohbot VD. Evidence aging on memory systems: contribution of two behavioral models in the mouse. for a virtual human analog of a rodent relational memory task: A study of aging and Curr Top Behav Neurosci (2012) 10:67–89. doi: 10.1007/7854_2011_151 fMRI in young adults. Hippocampus (2012) 22(4):869–80. doi: 10.1002/hipo.20948 45. Popovic D, Vieta E, Fornaro M, Perugi G. Cognitive tolerability following 31. Konishi K, Etchamendy N, Roy S, Marighetto A, Rajah N, Bohbot VD. successful long term treatment of major depression and anxiety disorders with SSRi Decreased functional magnetic resonance imaging activity in the hippocampus in antidepressants. J Affect Disord (2015) 173:211–5. doi: 10.1016/j.jad.2014.11.008 favor of the caudate nucleus in older adults tested in a virtual navigation task. 46. Bengtsson D, Ragnarsson O, Berinder K, Dahlqvist P, Eden Engstrom B, Hippocampus (2013) 23(11):1005–14. doi: 10.1002/hipo.22181 Ekman B, et al. Psychotropic drugs in patients with cushing’s disease before 32. Mingaud F, Mormede C, Etchamendy N, Mons N, Niedergang B, Wietrzych diagnosis and at long-term follow-up: a nationwide study. J Clin Endocrinol Metab M, et al. Retinoid hyposignaling contributes to aging-related decline in (2021) 106(6):1750–60. doi: 10.1210/clinem/dgab079 hippocampal function in short-term/working memory organization and long- 47. Papoian V, Biller BM, Webb SM, Campbell KK, Hodin RA, Phitayakorn R. term declarative memory encoding in mice. J Neurosci (2008) 28(1):279–91. doi: Patients’ perception on clinical outcome and quality of life after a diagnosis of 10.1523/JNEUROSCI.4065-07.2008 cushing syndrome. Endocr Pract (2016) 22(1):51–67. doi: 10.4158/EP15855.OR 33. Montgomery SA, Asberg M. A new depression scale designed to be sensitive 48. Badia X, Valassi E, Roset M, Webb SM. Disease-specific quality of life to change. Br J Psychiatry (1979) 134:382–9. doi: 10.1192/bjp.134.4.382 evaluation and its determinants in cushing’s syndrome: What have we learnt? 34. Snaith RP. The hospital anxiety and depression scale. Health Qual Life Pituitary (2014) 17(2):187–95. doi: 10.1007/s11102-013-0484-2 Outcomes (2003) 1:29. doi: 10.1186/1477-7525-1-29 49. Tiemensma J, Kaptein AA, Pereira AM, Smit JW, Romijn JA, Biermasz NR. 35. McHorney CA, Ware JEJr., Raczek AE. The MOS 36-item short-form health Negative illness perceptions are associated with impaired quality of life in patients survey (SF-36): II. psychometric and clinical tests of validity in measuring physical after long-term remission of cushing’s syndrome. Eur J Endocrinol (2011) 165 and mental health constructs. Med Care (1993) 31(3):247–63. doi: 10.1097/ (4):527–35. doi: 10.1530/EJE-11-0307 00005650-199303000-00006 50. Marinelli M, Piazza PV. Interaction between glucocorticoid hormones, 36. Webb SM, Badia X, Barahona MJ, Colao A, Strasburger CJ, Tabarin A, et al. stress and psychostimulant drugs. Eur J Neurosci (2002) 16(3):387–94. doi: Evaluation of health-related quality of life in patients with cushing’s syndrome with 10.1046/j.1460-9568.2002.02089.x a new questionnaire. Eur J Endocrinol (2008) 158(5):623–30. doi: 10.1530/EJE-07- 51. Piazza PV, Le Moal M. Glucocorticoids as a biological substrate of reward: 0762 physiological and pathophysiological implications. Brain Res Brain Res Rev (1997) 37. Marighetto A, Etchamendy N, Touzani K, Torrea CC, Yee BK, Rawlins JN, 25(3):359–72. doi: 10.1016/S0165-0173(97)00025-8 et al. Knowing which and knowing what: a potential mouse model for age-related 52. Vinson GP, Brennan CH. Addiction and the adrenal cortex. Endocrine human declarative memory decline. Eur J Neurosci (1999) 11(9):3312–22. doi: connections. (2013) 2(3):R1–R14. doi: 10.1530/EC-13-0028 10.1046/j.1460-9568.1999.00741.x 53. Koob GF. Neurobiology of opioid addiction: opponent process, 38. Sellami A, Abed ASA, Brayda-Bruno L, Etchamendy N, Valerio S, Oule M, hyperkatifeia, and negative reinforcement. Biol Psychiatry (2020) 87(1):44–53. et al. Protocols to study declarative memory formation in mice and humans : doi: 10.1016/j.biopsych.2019.05.023 optogenetics and translational behavioral approaches. Bio Protoc (2018) 8(12): e2888. doi: 10.21769/BioProtoc.2888 54. Nieman LK, Biller BM, Findling JW, Murad MH, Newell-Price J, Savage 39. Forget H, Lacroix A, Somma M, Cohen H. Cognitive decline in patients with MO, et al. Treatment of cushing’s syndrome: an endocrine society clinical practice cushing’s syndrome. J Int Neuropsychol Society: JINS (2000) 6(1):20–9. doi: guideline. J Clin Endocrinol Metab (2015) 100(8):2807–31. doi: 10.1210/jc.2015- 10.1017/S1355617700611037 1818 40. Psaras T, Milian M, Hattermann V, Gerlach C, Honegger J. Executive 55. Kelly JF, Greene MC, Bergman BG. Beyond abstinence: Changes in indices functions recover earlier than episodic memory after microsurgical of quality of life with time in recovery in a nationally representative sample of u.s. transsphenoidal resection of pituitary tumors in adult patients–a longitudinal adults. Alcohol Clin Exp Res (2018) 42(4):770–80. doi: 10.1111/acer.13604 study. J Clin Neurosci (2011) 18(10):1340–5. doi: 10.1016/j.jocn.2011.01.027 56. Pivonello R, Isidori AM, De Martino MC, Newell-Price J, Biller BM, Colao 41. Psaras T, Milian M, Hattermann V, Will BE, Tatagiba M, Honegger J. A. Complications of cushing’s syndrome: state of the art. Lancet Diabetes Predictive factors for neurocognitive function and quality of life after surgical Endocrinol (2016) 4(7):611–29. doi: 10.1016/S2213-8587(16)00086-3 Frontiers in Endocrinology 10 frontiersin.org