Prediction of outcomes by diffusion kurtosis imaging in patients with large (≥5 cm) hepatocellular carcinoma after liver resection: A retrospective study.
Front Oncol
; 12: 939358, 2022.
Article
em En
| MEDLINE
| ID: mdl-36465384
ABSTRACT
Purpose:
To evaluate preoperative diffusion kurtosis imaging (DKI) in predicting the outcomes of large hepatocellular carcinoma (HCC) after liver resection (LR). Materials andmethods:
From January 2015 to December 2017, patients with a large (≥5cm) HCC who underwent preoperative DKI were retrospectively reviewed. The correlations of the mean kurtosis (MK), mean diffusivity (MD), and apparent diffusion coefficient (ADC) with microvascular invasion (MVI) or histological grade were analyzed. Cox regression analyses were performed to identify the predictors of recurrence-free survival (RFS) and overall survival (OS). A nomogram to predict RFS was established. P<0.05 was considered as statistically significant.Results:
A total of 97 patients (59 males and 38 females, 56.0 ± 10.9 years) were included in this study. The MK, MD, and ADC values were correlated with MVI or histological grade (P<0.01). With a median follow-up time of 41.2 months (range 12-69 months), 67 patients (69.1%) experienced recurrence and 41 patients (42.3%) were still alive. The median RFS and OS periods after LR were 29 and 45 months, respectively. The 1-, 3-, and 5-year RFS and OS rates were 88.7%, 41.2%, and 21.7% and 99.0%, 68.3%, and 25.6%, respectively. MK (P<0.001), PVT (P<0.001), and ADC (P=0.033) were identified as independent predictor factors for RFS. A nomogram including the MK value for RFS showed the best performance, and the C-index was 0.895.Conclusion:
The MK value obtained from DKI is a potential predictive factor for recurrence and poor survival, which could provide valuable information for guiding the efficacy of LR in patients with large HCC.
Texto completo:
1
Base de dados:
MEDLINE
Tipo de estudo:
Observational_studies
/
Prognostic_studies
/
Risk_factors_studies
Idioma:
En
Ano de publicação:
2022
Tipo de documento:
Article