Almeida, João Paulo

Background: The complex nature and heterogeneity involving pituitary surgery results have increased interest in machine learning (ML) applications for prediction of outcomes over the last decade. This study aims to systematically review the characteristics of ML models involving pituitary surgery outcome prediction and assess their reporting quality. Methods: We searched the PubMed, Scopus, and Web of Knowledge databases for publications on the use of ML to predict pituitary surgery outcomes. We used the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) to assess report quality. Our search strategy was based on the terms “artificial intelligence”, “machine learning”, and “pituitary”. Results: 20 studies were included in this review. The principal models reported in each article were post-surgical endocrine outcomes (n = 10), tumor management (n = 3), and intra- and postoperative complications (n = 7). Overall, the included studies adhered to a median of 65% (IQR = 60–72%) of TRIPOD criteria, ranging from 43% to 83%. The median reported AUC was 0.84 (IQR = 0.80–0.91). The most popular algorithms were support vector machine (n = 5) and random forest (n = 5). Only two studies reported external validation and adherence to any reporting guideline. Calibration methods were not reported in 15 studies. No model achieved the phase of actual clinical applicability. Conclusion: Applications of ML in the prediction of pituitary outcomes are still nascent, as evidenced by the lack of any model validated for clinical practice. Although studies have demonstrated promising results, greater transparency in model development and reporting is needed to enable their use in clinical practice. Further adherence to reporting guidelines can help increase AI’s real-world utility and improve clinical practice.

AbstractBackgroundTraditional management of olfactory neuroblastoma (ONB) includes margin‐negative resection with removal of cribriform plate, dura, and olfactory bulb, regardless of intracranial disease. This approach may be overtreating certain patients. Our investigation examines risk factors associated with occult intracranial disease to optimize therapeutic outcomes.MethodsThis retrospective, multi‐institutional cohort study examined clinical covariates associated with occult intracranial involvement. Patient demographics, staging, Hyam's grade, and pathologic involvement of dura, olfactory bulb/tract, and brain were collected. Diagnostic imaging was reviewed. Positive and negative predictive value (NPV) were estimated along with effect size estimates. Cox hazard regression examined associations with overall survival (OS) and disease‐free survival (DFS).ResultsA total of 224 subjects with new diagnoses of ONB (2005–2021) were identified. Skull base bone involvement on computed tomography (CT) had the highest NPV for pathologic dura (88.0%), olfactory bulb (88%), and brain involvement (97.3%). Hyam's grade category was significantly associated with dural involvement (φC = 0.26; 95% confidence interval [CI]: 0.16, 0.42). Subjects without radiologic skull base involvement (n = 66) had pathologic positivity of 12.1%. Within this subgroup, Hyam's grade was clinically significant for dural positivity (φ = 0.34; 95% CI: −0.12, 0.71) with 28.6% involvement in high grade tumors. Neither clinical nor pathologic positivity of intracranial structures were associated with significantly different OS or DFS.ConclusionsBoth CT and magnetic resonance imaging (MRI) had reasonably good NPV for involvement of dura and olfactory bulb. Higher Hyam's grade was associated with dural involvement. Patients with low‐grade tumors not involving the skull base may be suitable for avoiding skull base resection; however, further investigation is warranted.

OBJECTIVE Surgery is the primary treatment for craniopharyngioma with the preservation of hypothalamic function of paramount importance. Infundibular preservation is debated, as maximal resection decreases recurrence rates but causes hypopituitarism. A triphasic response of diabetes insipidus (DI), syndrome of inappropriate antidiuretic hormone secretion (SIADH), and recurrent DI has been described after pituitary surgery, but the impact of infundibular preservation on the triphasic response following craniopharyngioma resection has not been well established. The authors’ objective was to assess postoperative fluid and sodium balance and differences in ADH imbalance management following endonasal craniopharyngioma resection based on infundibular transection status. METHODS This is a retrospective cohort study of 19 patients with craniopharyngioma treated with endoscopic endonasal resection between 2014 and 2021. Resection was dichotomized into infundibular transection or preservation. Postoperative triphasic response, time to DI, and time to ADH replacement were compared using Fisher’s exact test and Kaplan-Meier analysis. RESULTS Based on surgeon impression, 10 patients had infundibular transection and 9 had infundibular preservation. Overall, 16 patients experienced DI, 12 experienced persistent DI, and 6 experienced SIADH. A postoperative triphasic response occurred in 40% (n = 4) of transection patients without preoperative DI and 11% (n = 1) of preservation patients without preoperative DI. The median time to postoperative DI (0.5 vs 18.0 hours, p = 0.022) and median time to ADH replacement therapy (4.5 vs 24 hours, p = 0.0004) were significantly shorter in the transection group than in the preservation group. CONCLUSIONS Following endonasal craniopharyngioma resection, the triphasic response occurs in nearly half of infundibular transection cases. DI begins earlier with infundibular transection. On the basis of the study findings in which no patients met the criteria for SIADH or were endocrinologically unstable after postoperative day 6, it is reasonable to suggest that otherwise stable patients can be discharged at or before postoperative day 6 when ADH fluctuations have normalized and endocrinopathy is appropriately managed with oral desmopressin. Infundibular transection status may impact postoperative hormonal replacement strategies, but additional studies should evaluate their efficacies.

Introduction. Pituitary adenomas have the potential to infiltrate the dura mater, skull, and the venous sinuses. Tumor extension into the cavernous sinus is often observed in pituitary adenomas and techniques and results of surgery in this region are vastly discussed in the literature. Infiltration of parasellar dura and its impact for pituitary surgery outcomes is significantly less studied but recent studies have suggested a role of endoscopic resection of the medial wall of the cavernous sinus, in selected cases. In this study, we discuss the techniques and outcomes of recently proposed techniques for selective resection of the medial wall of the cavernous sinus in endoscopic pituitary surgery. Methods. We performed a systematic review of the literature using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and protocol and a total of 4 studies with 106 patients that underwent an endoscopic approach for resection of pituitary tumors with resection of medial wall from cavernous sinus were included. Clinical and radiological data were extracted (sex, mean age, Knosp, prior surgery, tumor size and type, complication rate, and remission) and a meta-analysis using the RevMan 5.4 software was performed. Results. A total of 5 studies with 208 patients were included in this analysis. The mean age of the study population was 48.87 years (range 25–82) with a female/male ratio of 1:1.36. Majority of the patients had Knosp Grade 1 (n = 77, 37.02%) and Grade 2 (n = 53, 25.48%). The complication rate was 4.81% (n = 33/106) and the most common complication observed was a new transient CN dysfunction and diplopia. Early disease remission was observed in 94.69% of the patients (n = 196/207). The prevalence rate of CS medial wall invasion varied from 10.4 % up to 36.7%. This invasion rate increased in frequency with higher Knosp Grade. The forest plot of persistent disease vs. remission in this surgery approach showed a p < 0.00001 and heterogeneity (I^2 = 0%). Discussion. Techniques to achieve resection of the medial wall of the cavernous sinus via the endoscopic endonasal approach include the “anterior to posterior” technique (opening of the anterior wall of the cavernous sinus) and the “medial to lateral” technique (opening of the inferior intercavernous sinus and). Although potentially related with improved endocrinological outcomes, these are advanced surgical techniques and require extensive anatomical knowledge and extensive surgical experience. Furthermore, to avoid procedure complications, extensive study of the patient’s configuration of cavernous ICA, Doppler-guided intraoperative imaging, surgical navigation system, and blunt tip knives to dissect the ICA’s plane are recommended. Conclusion. Endoscopic resection of the medial wall of the cavernous sinus has been associated with reports of high rates of postoperative hormonal control in functioning pituitary adenomas. However, it represents a more complex approach and requires advanced experience in endoscopic skull base surgery. Additional studies addressing case selection and studies evaluating long term results of this technique are still necessary.