Prediction of Recurrence after Transsphenoidal Surgery for Cushing’s Disease: The Use of Machine Learning Algorithms

2019 ◽  
Vol 108 (3) ◽  
pp. 201-210 ◽  
Author(s):  
Yifan Liu ◽  
Xiaohai Liu ◽  
Xinyu Hong ◽  
Penghao Liu ◽  
Xinjie Bao ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cushing’S Disease ◽  
Predictive Models ◽  
Transsphenoidal Surgery ◽  
Characteristic Curve ◽  
Serum Cortisol ◽  
Machine Learning Algorithms ◽  
Cushing's Disease ◽  
Morning Serum Cortisol ◽  
Better Than

Background: There are no reliable predictive models for recurrence after transsphenoidal surgery (TSS) for Cushing’s disease (CD). Objectives: This study aimed to develop machine learning (ML)-based predictive models for CD recurrence after initial TSS and to evaluate their performance. Method: A total of 354 CD patients were included in this retrospective, supervised learning, data mining study. Predictive models for recurrence were developed according to 17 variables using 7 algorithms. Models were evaluated based on the area under the receiver operating characteristic curve (AUC). Results: All patients were followed up for over 12 months (mean ± SD 43.80 ± 35.61). The recurrence rate was 13.0%. Age (p < 0.001), postoperative morning serum cortisol nadir (p = 0.002), and postoperative (p < 0.001) and preoperative (p = 0.04) morning adrenocorticotropin (ACTH) level were significantly related to recurrence. AUCs of the 7 models ranged from 0.608 to 0.781. The best performance (AUC = 0.781, 95% CI 0.706, 0.856) appeared when 8 variables were introduced to the random forest (RF) algorithm, which was much better than that of logistic regression (AUC = 0.684, p = 0.008) and that of using only postoperative morning serum cortisol (AUC = 0.635, p < 0.001). According to the feature selection algorithms, the top 3 predictors were age, postoperative serum cortisol, and postoperative ACTH. Conclusions: Using ML-based models for prediction of the recurrence after initial TSS for CD is feasible, and RF performs best. The performance of most of ML-based models was significantly better than that of some conventional models.

scholarly journals Surgical management of pediatric Cushing's disease: an analysis of 15 consecutive cases at a specialized neurosurgical center

2010 ◽  
Vol 54 (1) ◽  
pp. 17-23 ◽  
Author(s):  
Ricardo Santos de Oliveira ◽  
Margaret de Castro ◽  
Sonir Roberto Rauber Antonini ◽  
Carlos Eduardo Martinelli Júnior ◽  
Ayrton Custódio Moreira ◽  
...  
Keyword(s):  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Pediatric Patients ◽  
Serum Cortisol ◽  
Plasma Cortisol Level ◽  
Hormone Deficiency ◽  
Cushing's Disease ◽  
Specialized Center ◽  
Biochemical Cure

OBJECTIVE: The aim of this study was to review the results of surgery for pediatric patients with Cushing's disease who were less than 18 years old and underwent transsphenoidal surgery in a specialized center during a 25-year period. SUBJECTS AND METHODS: Retrospective study, in which the medical records, histology and pituitary imaging of 15 consecutive pediatric patients with Cushing's disease (mean age: 13 years) were evaluated by the same team of endocrinologists and a neurosurgeon from 1982 to 2006. Patients were considered cured when there was clinical adrenal insufficiency and serum cortisol levels were below 1. 8 µg/dL or 50 nmol/L after one, two, three, or seven days following surgery; they therefore required cortisone replacement therapy. Follow-up was for a median time of 11.5 years (range: 2 to 25 years). RESULTS: Clinical and biochemical cure was achieved in 9/15 patients (60%) exclusively after transsphenoidal surgery. Hypopituitarism was observed in four patients; growth hormone deficiency, in two; permanent diabetes insipidus, in one case. CONCLUSIONS: Cushing's disease is rare in children and adolescents. Transsphenoidal surgery is an effective and safe treatment in most of these patients. Plasma cortisol level < 1. 8 µg/dL following surgery is the treatment goal and is a good predictive factor for long-term cure of Cushing's disease.

scholarly journals Dynamics of Serum Cortisol Levels after Transsphenoidal Surgery in a Cohort of Patients with Cushing’s Disease

2004 ◽  
Vol 89 (3) ◽  
pp. 1131-1139 ◽  
Author(s):  
G. A. F. S. Rollin ◽  
N. P. Ferreira ◽  
M. Junges ◽  
J. L. Gross ◽  
M. A. Czepielewski
Keyword(s):  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Serum Cortisol ◽  
Cushing's Disease ◽  
Cortisol Levels

Evaluation of Cushing's disease remission after transsphenoidal surgery based on early serum cortisol dynamics

2013 ◽  
Vol 80 (3) ◽  
pp. 411-418 ◽  
Author(s):  
Fabíola Costenaro ◽  
Ticiana C. Rodrigues ◽  
Guilherme A. F. Rollin ◽  
Nelson P. Ferreira ◽  
Mauro A. Czepielewski
Keyword(s):  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Serum Cortisol ◽  
Cushing's Disease ◽  
Disease Remission

scholarly journals Morning Serum Cortisol as a Predictor for the HPA Axis Recovery in Cushing’s Disease

2021 ◽  
Vol 2021 ◽  
pp. 1-6
Author(s):  
Q. Cui ◽  
D. Liu ◽  
B. Xiang ◽  
Q. Sun ◽  
L. Fan ◽  
...  
Keyword(s):  
Hpa Axis ◽  
Cushing’S Disease ◽  
Serum Cortisol ◽  
Cross Sectional Study ◽  
Cushing's Disease ◽  
Confirmatory Test ◽  
Acth Stimulation ◽  
Cross Sectional ◽  
Postoperative Course ◽  
Morning Serum Cortisol

Background. The suppressed hypothalamic-pituitary-adrenal (HPA) axis after successful surgery for Cushing’s disease (CD) will recover in almost all patients. We aimed to identify the predictive factors for HPA axis recovery in CD patients with postoperative remission. Design and Methods. This observational retrospective cross-sectional study enrolled 69 CD patients with postoperative remission in Huashan Hospital from 2015 to 2019. All subjects had a detailed clinical evaluation. The low-dose ACTH stimulation test (LDT) was conducted as the gold standard for assessing the HPA axis function. Results. Peak cortisol in LDT was found only to be positively correlative with morning serum cortisol (MSC) ( ρ = 0.451 , p < 0.001 ). The MSC was higher ( p < 0.001 ), and the median postoperative course was significantly longer ( p = 0.025 ) in the patients with the recovered HPA axis function compared with unrecovered patients. The AUC value of MSC for predicting the recovery of the HPA axis was 0.701, and the optimal cutoff was 6.25 μg/dl (sensitivity 85.19% and specificity 47.62%). Other useful cutoff values were 10.74 μg/dl (specificity 100%) and 4.18 μg/dl (sensitivity 100%). Besides, combined with the postoperative course, the AUC values were higher than MSC alone (0.935 vs. 0.701, p < 0.001 ). Conclusions. MSC is a viable first-step diagnostic predictor for HPA axis recovery in CD patients with postoperative remission. For the patients with cortisol levels between 4.18 and 10.74 μg/dl, a confirmatory test should be conducted. When the MSC level was 10.74 μg/dl or greater, the replacement therapy could be discontinued.

scholarly journals An Ensemble Prediction Model for Potential Student Recommendation Using Machine Learning

Symmetry ◽  
2020 ◽  
Vol 12 (5) ◽  
pp. 728 ◽  
Author(s):  
Lijuan Yan ◽  
Yanshen Liu
Keyword(s):  
Machine Learning ◽  
Student Performance ◽  
Prediction Models ◽  
Characteristic Curve ◽  
Machine Learning Algorithms ◽  
Ensemble Prediction ◽  
Support Vector ◽  
Proposed Model ◽  
Importance Analysis ◽  
Better Than

Student performance prediction has become a hot research topic. Most of the existing prediction models are built by a machine learning method. They are interested in prediction accuracy but pay less attention to interpretability. We propose a stacking ensemble model to predict and analyze student performance in academic competition. In this model, student performance is classified into two symmetrical categorical classes. To improve accuracy, three machine learning algorithms, including support vector machine (SVM), random forest, and AdaBoost are established in the first level and then integrated by logistic regression via stacking. A feature importance analysis was applied to identify important variables. The experimental data were collected from four academic years in Hankou University. According to comparative studies on five evaluation metrics (precision, recall, F1, error, and area   under   the   receiver   operating   characteristic   curve ( AUC ) in this analysis, the proposed model generally performs better than compared models. The important variables identified from the analysis are interpretable, they can be used as guidance to select potential students.

scholarly journals Utility of the Immediate Postoperative Cortisol Concentrations in Patients With Cushing's Disease

Neurosurgery ◽  
2010 ◽  
Vol 67 (3) ◽  
pp. 688-695 ◽  
Author(s):  
Michael E. Sughrue ◽  
Jugal K. Shah ◽  
Jessica K. Devin ◽  
Sandeep Kunwar ◽  
Lewis S. Blevins
Keyword(s):  
Confidence Interval ◽  
Cortisol Level ◽  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Fixed Effects ◽  
Serum Cortisol ◽  
Cushing's Disease ◽  
Recurrence Rates ◽  
Cumulative Rate ◽  
Cortisol Levels

Abstract BACKGROUND Several investigators have recommended serial measurements of serum cortisol in the days following pituitary surgery to identify patients at risk of recurrence. OBJECTIVE We systematically reviewed the literature on this topic and analyzed the usefulness of this test in our own patient population. METHODS We identified studies publishing data regarding recurrence rates after transsphenoidal surgery for Cushing's disease, focusing on studies with data regarding patients with early postoperative cortisol levels. We determined a cumulative relative risk of having a subnormal vs normal cortisol level postoperatively using a fixed-effects meta-analysis model. Additionally, we analyzed our own patients with Cushing's disease undergoing transsphenoidal surgery and performed Kaplan-Meier analysis of recurrence-free survival for patients with undetectable, subnormal but detectable, and normal immediate 8 AM serum cortisol levels. RESULTS Fourteen studies met inclusion criteria. The length of follow-up varied between 32 and 115 months. The cumulative rate of recurrence in the group of patients with subnormal cortisol levels was 9% (95% confidence interval: 6%–12%). The cumulative rate of recurrence in the group with normal cortisol levels was 24% (95% confidence interval: 17%–31%). We analyzed 73 of our own patients and found similar recurrence rates in patients with subnormal vs normal early postoperative cortisol levels (4% vs 22%, χ2 test, P &lt; .05). CONCLUSION Although a subnormal early postoperative cortisol level is predictive of improved outcome after transsphenoidal surgery for Cushing's disease, it is not analogous with cure, nor is a normal level completely predictive of future failure.

scholarly journals Comparisons in the epidemiology, diagnostic features and cure rate by transsphenoidal surgery between paediatric and adult-onset Cushing's disease

2011 ◽  
Vol 164 (5) ◽  
pp. 667-674 ◽  
Author(s):  
H L Storr ◽  
K I Alexandraki ◽  
L Martin ◽  
A M Isidori ◽  
G A Kaltsas ◽  
...  
Keyword(s):  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Clinical Presentation ◽  
Serum Cortisol ◽  
University Hospital ◽  
Cushing's Disease ◽  
Adult Onset ◽  
Cure Rate ◽  
Diagnostic Features ◽  
Magnetic Resonance Imaging Mri

ObjectiveThere are few published comparisons between paediatric and adult-onset Cushing's disease (CD). We compare the epidemiology, diagnostic features and cure rate by transsphenoidal surgery (TSS) in these groups.DesignRetrospective review of patient databases in a single university hospital centre.PatientsTotally, 41 paediatric (mean age 12.3±3.5 years; range 5.7–17.8) and 183 adult (mean age 40±13 years; range 18.0–95.0) patients with CD were investigated.ResultsPaediatric CD was characterised by male (63%) and adult CD by a female predominance (79%, P<0.0001). There were small but significant differences in clinical presentation. Biochemical features of CD were comparable except the serum cortisol increase during a CRH test: mean change (105%, n=39) in paediatric and (54%, n=123) in adult subjects (P<0.0001). Macroadenomas were more common in adult (15%, 28/183) than in paediatric (2%, 1/41, P=0.04) CD. Corticotroph microadenomas were more easily visualised by pituitary magnetic resonance imaging (MRI) in adult (76%, 50/66) compared with paediatric (55%, 21/38, P=0.045) CD with poorer concordance of imaging with surgical findings in children (P=0.058). The incidence of ACTH lateralisation by bilateral simultaneous inferior petrosal sinus sampling was comparable in paediatric (76%, 25/33) and adult (79%, 46/58; P=0.95) patients with good surgical concordance in both (82% paediatric and 79% adult). Cure rates by TSS were comparable, with a paediatric cure rate of 69%.ConclusionSeveral features of paediatric CD are distinct: increased frequency of prepubertal CD in males, the different clinical presentation, the decreased presence of macroadenomas and the frequent absence of radiological evidence of an adenoma on MRI.

scholarly journals Prediction of the transsphenoidal endoscopic adenomectomy results in patients with cushing’s disease

2022 ◽  
pp. 152-161
Author(s):  
N. V. Kuritsyna ◽  
U. A. Tsoy ◽  
V. Yu. Cherebillo ◽  
A. A. Paltsev ◽  
A. I. Tsiberkin ◽  
...  
Keyword(s):  
Cushing’S Disease ◽  
Serum Cortisol ◽  
Cushing's Disease ◽  
Serum Cortisol Level ◽  
High Dose ◽  
High Dose Dexamethasone ◽  
Morning Serum Cortisol ◽  
Transsphenoidal Adenomectomy ◽  
Important Challenge ◽  
One Year

Introduction. Transsphenoidal adenomectomy (TSA) is the method of choice in the treatment of Cushing’s disease (CD), but remission of hypercorticism cannot be achieved in all patients. The search for predictors of CD remission after TSA remains to be an important challenge in the endocrinology today.Aim. To study the preoperative and postoperative data of patients with CD to identify the predictors of hypercorticism remission after TSA.Materials and methods. 101 patients with confirmed CD after TSA were included. One year after operation all patients were examined for the presence of hypercorticism remission and divided into two groups: with CD remission and its absence. In both groups’ preoperative pituitary magnetic resonance imaging (MRI) data, the results of preoperative high dose dexamethasone suppression test (HDDST) and the results of serum cortisol collected in the morning 2–3 days after surgery were compared.Results. One year after TSA, CD remission was confirmed in 63 patients (62.4%), whereas in 38 patients (37.6%) hypercortisolism persisted. Favorable predictors of CD remission were: the adenoma size > 3 mm without the invasive growth according to pituitary MRI (specificity 82.4%, sensitivity 82.8%), serum cortisol suppression ≥ 74% in preoperative HDDST (specificity 81.5%, sensitivity 86.3%), morning serum cortisol level in 2–3 days after surgery ≤ 388 nmol/l (specificity 79.3%, sensitivity 97.4%).Conclusions. Pituitary MRI data, the results of preoperative HDDST and morning serum cortisol in the 2–3 days after surgery can be used as predictors of CD remission.

Diagnosis and selective cure of Cushing's disease during pregnancy by transsphenoidal surgery

1995 ◽  
Vol 132 (6) ◽  
pp. 722-726 ◽  
Author(s):  
RJM Ross ◽  
SL Chew ◽  
L Perry ◽  
K Erskine ◽  
S Medbak ◽  
...  
Keyword(s):  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Serum Cortisol ◽  
Cushing's Disease ◽  
Successful Outcome ◽  
Small Female ◽  
Free Cortisol ◽  
Cortisol Levels ◽  
Biochemical Features ◽  
Mr Scanning

Ross RJM, Chew SL, Perry L, Erskine K, Medbak S, Afshar F. Diagnosis and selective cure of Cushing's disease during pregnancy by transsphenoidal surgery. Eur J Endocrinol 1995;132:722–6. ISSN 0804–4643 The diagnosis of Cushing's Disease during pregnancy is complex because the biochemical features are obscured by changes in the normal hypothalamo-pituitary-adrenal axis that occur during gestation. To date, treatment has not been successful and there is a high incidence of maternal and fetal complications. We report the case of a 24-year-old woman with Cushing's disease who presented during her 16th week of pregnancy. Diagnosis was confirmed by the finding of elevated serum and urinary free cortisol levels with loss of the normal circadian rhythm of serum cortisol. Cortisol levels failed to suppress after a low-dose dexamethasone test but suppressed after a high-dose test. There was an exaggerated serum cortisol and plasma adrenocorticotrophin (ACTH) response to corticotrophinreleasing hormone (CRH). Magnetic resonance (MR) scanning demonstrated a pituitary tumour and cure was effected by transsphenoidal surgery where tumour immunostaining for ACTH was removed. Postoperatively the patient made an uncomplicated recovery; serum cortisol and plasma ACTH levels were undetectable at 9 days following surgery and recovery of the hypothalamo-pituitary axis occurred at 99 days after surgery. Caesarean section was performed at 38 weeks of pregnancy and a healthy but small female infant was delivered. This case illustrates the biochemical features of Cushing's disease during pregnancy and is the first report of the use of CRH testing and MR scanning in this clinical situation. The cure by surgery and successful outcome for mother and infant, with preservation of normal anterior pituitary function, suggest that transsphenoidal surgery may be the treatment of choice. RJM Ross, Dept of Endocrinology, St Bartholomew's Hospital, London ECIA 7BE, UK

scholarly journals Dynamics of postoperative serum cortisol after transsphenoidal surgery for Cushing’s disease: implications for immediate reoperation and remission

2018 ◽  
Vol 129 (5) ◽  
pp. 1268-1277 ◽  
Author(s):  
Marc Mayberg ◽  
Stephen Reintjes ◽  
Anika Patel ◽  
Kelley Moloney ◽  
Jennifer Mercado ◽  
...  
Keyword(s):  
Cushing’S Disease ◽  
Transsphenoidal Surgery ◽  
Serum Cortisol ◽  
Late Recurrence ◽  
Cushing's Disease ◽  
Entire Cohort ◽  
Temporal Course ◽  
Postoperative Serum ◽  
Cortisol Levels ◽  
Immediate Reoperation

OBJECTIVESuccessful transsphenoidal surgery for adrenocorticotropin hormone (ACTH)–producing pituitary tumors is associated with subnormal postoperative serum cortisol levels, which may guide decisions regarding immediate reoperation. However, little is known about the detailed temporal course of changes in serum cortisol in the immediate postoperative period, and the relationship of postoperative cortisol dynamics to remission and late recurrence.METHODSA single-center retrospective cohort analysis was performed for all patients undergoing pituitary surgery from 2007 through 2015. Standardized diagnostic and treatment algorithms were applied to all patients with potential Cushing’s disease (CD), including microsurgical transsphenoidal adenomectomy (TSA) by a single surgeon. All patients had serum cortisol levels drawn at 6-hour intervals for 72 hours after surgery, and were offered reoperation within 3 days for normal or supranormal postoperative cortisol levels. Primary outcomes were 6-month remission and late recurrence; secondary outcomes were persistent postoperative hypocortisolism and surgical morbidity. Discriminatory levels of postoperative serum cortisol for predicting remission were calculated at various intervals after surgery using receiver operating characteristic (ROC) curves.RESULTSAmong 89 patients diagnosed with CD, 81 underwent initial TSA for a potentially curable lesion; 23 patients (25.8%) underwent an immediate second TSA. For the entire cohort, 6-month remission was achieved in 77.8% and late recurrences occurred in 9.5%, at a mean of 43.5 months. Compared with patients with a single surgery, those with an immediate second TSA had similar rates of remission (78.3% vs 77.6%) and late recurrence (5.6% vs 11.1%). The rate of hypocortisolism for patients with 2 surgeries (12/23, 52.2%) was significantly greater than that for patients with single surgeries (13/58, 22.4%; p < 0.001). There was no difference in the incidence of CSF leaks between the first and second operations. Remission was achieved in 58 (92.1%) of 64 patients who completed the 2-surgery protocol. The temporal course of postoperative serum cortisol levels among patients varied considerably, with subnormal nadir levels < 2 μg/dl occurring between 12 hours and 66 hours. Patients achieving remission had significantly lower mean serum cortisol levels at every time point after surgery (p < 0.01). By ROC curve analysis, nadir cortisol levels < 2.1 μg/dl were predictive of 6-month remission for the entire cohort over 3 days (positive predictive value [PPV] = 94%); discriminating cortisol levels for predicting remission on postoperative day (POD) 2 were < 5.4 μg/dl (PPV = 97%), although patients with remission after postoperative cortisol levels of 2–5 μg/dl had a significantly higher rate of late recurrence.CONCLUSIONSThere is substantial variation in the temporal course of serum cortisol levels over the first 72 hours after TSA for CD, with nadir levels predictive for remission occurring as late as POD 3. Although a cortisol level of 2.1 μg/dl at any point was an accurate predictor of 6-month remission, levels less than 5.4 μg/dl on POD 2 were reasonably accurate. These data may enable decisions regarding the efficacy of an immediate second surgical procedure performed during the same hospitalization; immediate reoperation is associated with excellent remission rates and low recurrence rates in patients otherwise unlikely to achieve remission, but carries a higher risk of permanent hypocortisolism.

Sign in / Sign up

Export Citation Format

Share Document