The prognosis of patients undergoing transarterial chemoembolization (TACE) is extremely variable, and a confounding factor is that TACE is often repeated several times. We retrospectively evaluated the accuracy of different prognostic scores and staging systems in estimating overall survival (OS) in patients with hepatocellular carcinoma (HCC). Methods: An analysis considering prognostic models as time-varying variables was performed, calculating OS from the time of TACE to the time of the subsequent treatment. Total follow-up time for each patient was therefore split into several observation times accounting for each TACE procedure. Values of the likelihood ratio test (LRT) and Akaike information criterion (AIC) were used to compare different systems. Univariable and multivariable analyses were conducted to identify additional factors predictive of OS. We analyzed 1,610 TACE performed in 1,058 patients recorded in the Italian Liver Cancer database from 2008 through 2016. Results: The median OS of the enrolled patients was 41 months. According to LRT χ2 and AIC values based on the time-varying analysis, mHAP-III achieved the best values (41.72 and 4,625.49, respectively, p < 0.0001), indicating the highest predictive performance compared with all other scores (HAP, mHAP-II, ALBI, and pALBI) and staging systems (MELD, ITALICA, CLIP, MESH, MESIAH, JIS, HKLC, and BCLC). In the multivariable Cox proportional hazards model, mHAP-III maintained an independent effect on OS (hazard ratio 1.31, 95% CI: 1.10-1.55, p < 0.0001). Time-varying age, alcoholic etiology, radiologic response to TACE, and performing ablation or surgery after TACE were additional significant variables resulting from the multivariable model. Conclusion: An innovative time-varying analysis revealed that mHAP-III was the most accurate model in predicting OS in patients with HCC undergoing TACE. Other clinical pre- and post-TACE variables were also found to be relevant for this prediction.

Time-Varying mHAP-III Is the Most Accurate Predictor of Survival in Patients with Hepatocellular Carcinoma Undergoing Transarterial Chemoembolization

Tovoli F.;Zoli M.;Azzaroli F.;Trevisani F.;
2021

Abstract

The prognosis of patients undergoing transarterial chemoembolization (TACE) is extremely variable, and a confounding factor is that TACE is often repeated several times. We retrospectively evaluated the accuracy of different prognostic scores and staging systems in estimating overall survival (OS) in patients with hepatocellular carcinoma (HCC). Methods: An analysis considering prognostic models as time-varying variables was performed, calculating OS from the time of TACE to the time of the subsequent treatment. Total follow-up time for each patient was therefore split into several observation times accounting for each TACE procedure. Values of the likelihood ratio test (LRT) and Akaike information criterion (AIC) were used to compare different systems. Univariable and multivariable analyses were conducted to identify additional factors predictive of OS. We analyzed 1,610 TACE performed in 1,058 patients recorded in the Italian Liver Cancer database from 2008 through 2016. Results: The median OS of the enrolled patients was 41 months. According to LRT χ2 and AIC values based on the time-varying analysis, mHAP-III achieved the best values (41.72 and 4,625.49, respectively, p < 0.0001), indicating the highest predictive performance compared with all other scores (HAP, mHAP-II, ALBI, and pALBI) and staging systems (MELD, ITALICA, CLIP, MESH, MESIAH, JIS, HKLC, and BCLC). In the multivariable Cox proportional hazards model, mHAP-III maintained an independent effect on OS (hazard ratio 1.31, 95% CI: 1.10-1.55, p < 0.0001). Time-varying age, alcoholic etiology, radiologic response to TACE, and performing ablation or surgery after TACE were additional significant variables resulting from the multivariable model. Conclusion: An innovative time-varying analysis revealed that mHAP-III was the most accurate model in predicting OS in patients with HCC undergoing TACE. Other clinical pre- and post-TACE variables were also found to be relevant for this prediction.
2021
Campani C.; Vitale A.; Dragoni G.; Arena U.; Laffi G.; Cillo U.; Giannini E.G.; Tovoli F.; Rapaccini G.L.; Di Marco M.; Caturelli E.; Zoli M.; Sacco R.; Cabibbo G.; Mega A.; Guarino M.; Gasbarrini A.; Svegliati-Baroni G.; Foschi F.G.; Biasini E.; Masotto A.; Nardone G.; Raimondo G.; Azzaroli F.; Vidili G.; Brunetto M.R.; Farinati F.; Trevisani F.; Marra F.
File in questo prodotto:
File Dimensione Formato  
lic-0010-0126.pdf

accesso aperto

Tipo: Versione (PDF) editoriale
Licenza: Licenza per Accesso Aperto. Creative Commons Attribuzione - Non commerciale - Non opere derivate (CCBYNCND)
Dimensione 384.31 kB
Formato Adobe PDF
384.31 kB Adobe PDF Visualizza/Apri
supplementary materials.zip

accesso aperto

Tipo: File Supplementare
Licenza: Licenza per accesso libero gratuito
Dimensione 52.66 kB
Formato Zip File
52.66 kB Zip File Visualizza/Apri

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/11585/873207
Citazioni
  • ???jsp.display-item.citation.pmc??? 3
  • Scopus 4
  • ???jsp.display-item.citation.isi??? 6
social impact