Purpose: To investigate the role of CT-texture analysis of liver ablation area to predict local recurrence after HCC ablation. Methods: Patients treated with liver ablation were retrospectively enrolled. CT-texture analysis was performed on the core and borders of ablation area 1-2 months after procedure. Tumors were grouped according to the onset of local recurrence at follow-up (persistence, recurrence-free, short- or long-term recurrence). Differences in texture parameters and which parameters were predictive of recurrence risk were assessed using a Cox regression model. Results: 151 HCCs were treated in 98 patients (72 ± 9 years, 83 men). 68 HCCs reported no disease recurrence, 32 persistent disease, 19 short-term and 32 long-term recurrence. Median follow-up was 280 [IQR: 156-569] days. Venous phase (Ven)Skewness (HR 6.07, 1.29-28.6, p =.02) and VenKurtosis (HR 2.27, 1.23-4.21, p =.01) of the ablation core were predictive of short-term recurrence. VenHUmean (HR 0.30, 0.11-0.81, p =.02) and VenGLRLM_HGRE (HR 1.06, 1.01-1.11, p =.02) of the core were independent predictors of tumor recurrence (C-index 0.64, CI 0.52-0.76, p =.03). Arterial phase (Art)Entropy of ablation border predicted the recurrence risk (HR 3.15, 1.05-9.42, p =.04) and values higher than 3.71 reported an increased recurrence incidence (p =.05). ArtHUstd (HR 1.14, 1.04-1.24, p =.01), LateHUmean (HR 8.69, 1.11-68.23, p =.04), LateGLRLM_HGRE (HR 0.9, 0.82-0.99, p =.03), LateGLZLM_HGZE (HR 1.01, 1.00-1.02, p <.01) and LateGLZLM_SZHGE (HR 0.99, 0.99-1.00, p =.02) of ablation border were independent predictors of local recurrence risk (C-index 0.73, CI 0.61-0.86, p <.01). Conclusions: CT texture analysis of ablation area performed at 1-2 months follow-up could estimate the risk of local recurrence of hepatocellular carcinoma treated by radiofrequency ablation.

Radiofrequency ablation of hepatocellular carcinoma: CT texture analysis of the ablated area to predict local recurrence

Beleù, Alessandro;Autelitano, Daniele;Geraci, Luca;Aluffi, Gregorio;Cardobi, Nicolò;De Robertis, Riccardo;Martone, Enrico;Conci, Simone;Ruzzenente, Andrea;D'Onofrio, Mirko
2022-01-01

Abstract

Purpose: To investigate the role of CT-texture analysis of liver ablation area to predict local recurrence after HCC ablation. Methods: Patients treated with liver ablation were retrospectively enrolled. CT-texture analysis was performed on the core and borders of ablation area 1-2 months after procedure. Tumors were grouped according to the onset of local recurrence at follow-up (persistence, recurrence-free, short- or long-term recurrence). Differences in texture parameters and which parameters were predictive of recurrence risk were assessed using a Cox regression model. Results: 151 HCCs were treated in 98 patients (72 ± 9 years, 83 men). 68 HCCs reported no disease recurrence, 32 persistent disease, 19 short-term and 32 long-term recurrence. Median follow-up was 280 [IQR: 156-569] days. Venous phase (Ven)Skewness (HR 6.07, 1.29-28.6, p =.02) and VenKurtosis (HR 2.27, 1.23-4.21, p =.01) of the ablation core were predictive of short-term recurrence. VenHUmean (HR 0.30, 0.11-0.81, p =.02) and VenGLRLM_HGRE (HR 1.06, 1.01-1.11, p =.02) of the core were independent predictors of tumor recurrence (C-index 0.64, CI 0.52-0.76, p =.03). Arterial phase (Art)Entropy of ablation border predicted the recurrence risk (HR 3.15, 1.05-9.42, p =.04) and values higher than 3.71 reported an increased recurrence incidence (p =.05). ArtHUstd (HR 1.14, 1.04-1.24, p =.01), LateHUmean (HR 8.69, 1.11-68.23, p =.04), LateGLRLM_HGRE (HR 0.9, 0.82-0.99, p =.03), LateGLZLM_HGZE (HR 1.01, 1.00-1.02, p <.01) and LateGLZLM_SZHGE (HR 0.99, 0.99-1.00, p =.02) of ablation border were independent predictors of local recurrence risk (C-index 0.73, CI 0.61-0.86, p <.01). Conclusions: CT texture analysis of ablation area performed at 1-2 months follow-up could estimate the risk of local recurrence of hepatocellular carcinoma treated by radiofrequency ablation.
Hepatocellular carcinoma, Local recurrence, Radiofrequency ablation, Radiomics, Texture analysis
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11562/1061275
Citazioni
  • ???jsp.display-item.citation.pmc??? 0
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 0
social impact