Frontiers in Oncology (May 2022)

Myosteatosis can Predict Unfavorable Outcomes in Advanced Hepatocellular Carcinoma Patients Treated With Hepatic Artery Infusion Chemotherapy and Anti-PD-1 Immunotherapy

  • Xiaoping Yi,
  • Xiaoping Yi,
  • Xiaoping Yi,
  • Xiaoping Yi,
  • Xiaoping Yi,
  • Yan Fu,
  • Yan Fu,
  • Qianyan Long,
  • Qianyan Long,
  • Yazhuo Zhao,
  • Yazhuo Zhao,
  • Sai Li,
  • Sai Li,
  • Chunhui Zhou,
  • Chunhui Zhou,
  • Huashan Lin,
  • Xiaolian Liu,
  • Chang Liu,
  • Changyong Chen,
  • Liangrong Shi,
  • Liangrong Shi

DOI
https://doi.org/10.3389/fonc.2022.892192
Journal volume & issue
Vol. 12

Abstract

Read online

AimTo evaluate the feasibility of computed tomography (CT) - derived measurements of body composition parameters to predict the risk factor of non-objective response (non-OR) in patients with hepatocellular carcinoma (HCC) undergoing anti-PD-1 immunotherapy and hepatic artery infusion chemotherapy (immune-HAIC).MethodsPatients with histologically confirmed HCC and treated with the immune-HAIC were retrospectively recruited between June 30, 2019, and July 31, 2021. CT-based estimations of body composition parameters were acquired from the baseline unenhanced abdominal CT images at the level of the third lumbar vertebra (L3) and were applied to develop models predicting the probability of OR. A myosteatosis nomogram was built using the multivariate logistic regression incorporating both myosteatosis measurements and clinical variables. Receiver operating characteristic (ROC) curves assessed the performance of prediction models, including the area under the curve (AUC). The nomogram’s performance was assessed by the calibration, discrimination, and decision curve analyses. Associations among predictors and gene mutations were also examined by correlation matrix analysis.ResultsFifty-two patients were recruited to this study cohort, with 30 patients having a OR status after immune-HAIC treatment. Estimations of myosteatosis parameters, like SM-RA (skeletal muscle radiation attenuation), were significantly associated with the probability of predicting OR (P=0.007). The SM-RA combined nomogram model, including serum red blood cell, hemoglobin, creatinine, and the mean CT value of visceral fat (VFmean) improved the prediction probability for OR disease with an AUC of 0.713 (95% CI, 0.75 to 0.95) than the clinical model nomogram with AUC of 0.62 using a 5-fold cross-validation methodology. Favorable clinical potentials were observed in the decision curve analysis.ConclusionsThe CT-based estimations of myosteatosis could be used as an indicator to predict a higher risk of transition to the Non-OR disease state in HCC patients treated with immune-HAIC therapy. This study demonstrated the therapeutic relevance of skeletal muscle composition assessments in the overall prediction of treatment response and prognosis in HCC patients.

Keywords