Shi-Wei Li1,2, Jing Zhang1, Han-Lu Tang2, Peng Li2, Bo Wang2, Fu Zhao3,4, Pi-Nan Liu5,6. 1. Department of Neural Reconstruction, Beijing Key Laboratory of Central Nervous System Injury, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China. 2. Neurosurgical Department, Beijing TianTan Hospital, Capital Medical University, Beijing, China. 3. Department of Neural Reconstruction, Beijing Key Laboratory of Central Nervous System Injury, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China. zhaofu@ccmu.edu.cn. 4. Neurosurgical Department, Beijing TianTan Hospital, Capital Medical University, Beijing, China. zhaofu@ccmu.edu.cn. 5. Department of Neural Reconstruction, Beijing Key Laboratory of Central Nervous System Injury, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China. pinanliu@ccmu.edu.cn. 6. Neurosurgical Department, Beijing TianTan Hospital, Capital Medical University, Beijing, China. pinanliu@ccmu.edu.cn.
Abstract
INTRODUCTION: Treatment for vestibular schwannoma (VS) in patients with neurofibromatosis type 2 (NF2) is extremely challenging due to the high risk of hearing loss. The aim of this study was to develop nomograms for the prediction of useful hearing loss in patients with NF2. METHODS: The nomogram was based on a retrospective study of 111 NF2 patients who underwent resection of large VS (> 2 cm) at Beijing Tiantan Hospital between 2011 and 2018. The utility of the proposed nomogram models was evaluated by receiver operating characteristic (ROC) curve, area under ROC curve (AUC), and calibration curve. The results were validated using a prospective cohort study on 33 patients consecutively enrolled at the same institution from 2019 to 2021. RESULTS: On multivariate analysis of the primary cohort, large tumour size (> 3 cm) and long duration of symptoms (> 24 months) were independent risk factors for preoperative useful hearing loss (AAO-HNS Class D) (P = 0.001 and P = 0.011, respectively), while large tumour size (> 3 cm), poor hearing (Class C), and lobular growth were significantly related to postoperative useful hearing loss (P < 0.001, P = 0.031 and P = 0.033, respectively). Factors derived from multivariable analysis were all assembled into the nomogram. The calibration curve for probability of hearing loss showed good agreement between predictions by nomogram models and actual observation. The ROC curves showed good predictive accuracy of the nomogram models in both cohorts (AUC: 0.708 to 0.951). CONCLUSION: The proposed nomograms resulted in accurate predictions of hearing outcomes for patients with NF2.
INTRODUCTION: Treatment for vestibular schwannoma (VS) in patients with neurofibromatosis type 2 (NF2) is extremely challenging due to the high risk of hearing loss. The aim of this study was to develop nomograms for the prediction of useful hearing loss in patients with NF2. METHODS: The nomogram was based on a retrospective study of 111 NF2 patients who underwent resection of large VS (> 2 cm) at Beijing Tiantan Hospital between 2011 and 2018. The utility of the proposed nomogram models was evaluated by receiver operating characteristic (ROC) curve, area under ROC curve (AUC), and calibration curve. The results were validated using a prospective cohort study on 33 patients consecutively enrolled at the same institution from 2019 to 2021. RESULTS: On multivariate analysis of the primary cohort, large tumour size (> 3 cm) and long duration of symptoms (> 24 months) were independent risk factors for preoperative useful hearing loss (AAO-HNS Class D) (P = 0.001 and P = 0.011, respectively), while large tumour size (> 3 cm), poor hearing (Class C), and lobular growth were significantly related to postoperative useful hearing loss (P < 0.001, P = 0.031 and P = 0.033, respectively). Factors derived from multivariable analysis were all assembled into the nomogram. The calibration curve for probability of hearing loss showed good agreement between predictions by nomogram models and actual observation. The ROC curves showed good predictive accuracy of the nomogram models in both cohorts (AUC: 0.708 to 0.951). CONCLUSION: The proposed nomograms resulted in accurate predictions of hearing outcomes for patients with NF2.
Authors: Jaishri O Blakeley; D Gareth Evans; John Adler; Derald Brackmann; Ruihong Chen; Rosalie E Ferner; C Oliver Hanemann; Gordon Harris; Susan M Huson; Abraham Jacob; Michel Kalamarides; Matthias A Karajannis; Bruce R Korf; Victor-Felix Mautner; Andrea I McClatchey; Harry Miao; Scott R Plotkin; William Slattery; Anat O Stemmer-Rachamimov; D Bradley Welling; Patrick Y Wen; Brigitte Widemann; Kim Hunter-Schaedle; Marco Giovannini Journal: Am J Med Genet A Date: 2011-12-02 Impact factor: 2.802
Authors: Ashok R Asthagiri; Dilys M Parry; John A Butman; H Jeffrey Kim; Ekaterini T Tsilou; Zhengping Zhuang; Russell R Lonser Journal: Lancet Date: 2009-05-22 Impact factor: 79.321
Authors: William H Slattery; Michael Hoa; Nicolas Bonne; Rick A Friedman; Marc S Schwartz; Laurel M Fisher; Derald E Brackmann Journal: Otol Neurotol Date: 2011-08 Impact factor: 2.311
Authors: Carlotta Morselli; N Boari; M Artico; M Bailo; L O Piccioni; I Giallini; M de Vincentiis; P Mortini; P Mancini Journal: Neurosurg Rev Date: 2020-02-06 Impact factor: 3.042
Authors: Michael S Dirks; John A Butman; H Jeffrey Kim; Tianxia Wu; Keaton Morgan; Anne P Tran; Russell R Lonser; Ashok R Asthagiri Journal: J Neurosurg Date: 2012-04-13 Impact factor: 5.115
Authors: James R Tysome; Robert Macfarlane; Juliette Durie-Gair; Neil Donnelly; Richard Mannion; Richard Knight; Frances Harris; Zebunnisa H Vanat; Yu Chen Tam; Katherine Burton; Anke Hensiek; F Lucy Raymond; David A Moffat; Patrick R Axon Journal: Otol Neurotol Date: 2012-04 Impact factor: 2.311
Authors: Roland Goldbrunner; Michael Weller; Jean Regis; Morten Lund-Johansen; Pantelis Stavrinou; David Reuss; D Gareth Evans; Florence Lefranc; Kita Sallabanda; Andrea Falini; Patrick Axon; Olivier Sterkers; Laura Fariselli; Wolfgang Wick; Joerg-Christian Tonn Journal: Neuro Oncol Date: 2020-01-11 Impact factor: 12.300