RESUMO
BACKGROUND: The factors influencing relapse after radiofrequency operation of the V2 branch of the trigeminal neuralgia are yet to be identified. OBJECTIVES: The risk factors affecting recurrence after radiofrequency operation of the V2 branch of the trigeminal neuralgia were analyzed, and a curative effect prediction model was constructed. STUDY DESIGN: A retrospective study. SETTING: This study was conducted at the Affiliated Hospital of Jiaxing University, People's Republic of China. METHODS: The records of patients with maxillary nerve pain in the V2 branch of the trigeminal nerve who underwent computed tomography-guided foramen rotundum radiofrequency treatment at the Pain Department of the Affiliated Hospital of Jiaxing College from April 2014 through December 2020 were collected and randomly divided into training (n = 137) and test (n = 59) groups at a 7:3 ratio. The outcome variable was whether or not recurrence was observed 2 years postsurgery. Independent predictors were screened by LASSO (least absolute shrinkage and selection operator) regression analysis. Based on these findings, a nomogram prediction model was explored further and developed using multifactor logistic regression analysis. Also, the feasibility of the nomogram prediction model for recurrence after radiofrequency was assessed using a validation group. Finally, the discriminatory power, accuracy, and clinical utility of the prediction model were evaluated using the receiver operating characteristic (ROC), area under the curve (AUC), calibration curve, and decision curve analysis (DCA), respectively. RESULTS: LASSO regression, combined with multifactorial logistic regression analysis, identified factors such as age, duration, branches, and numbness that influence V2 trigeminal nerve pain recurrence in patients 2 years post-radiofrequency surgery (P < 0.05). The above variables were used to construct the nomogram prediction models. The AUC of the nomogram prediction model predicted that the recurrence post V2 radiofrequency was 0.726 in the training group and 0.611 in the test group. The DCA showed that the columnar plot prediction model predicted the risk of recurrence post-radiofrequency of the V2 branch of the trigeminal nerve had a threshold probability of 0 - 0.9. LIMITATIONS: This was a single-center study. CONCLUSION: A highly accurate nomogram prediction model (predictor variables include age, duration, branches, and numbness) was developed to improve the early identification and screening of patients at high risk of recurrence after V2 trigeminal nerve radiofrequency surgery. KEY WORDS: Trigeminal neuralgia, maxillary neuralgia, radiofrequency, risk factors, prediction model.
