Inoperable hepatocellular carcinoma treated with transcatheter arterial chemoembolisation (TACE): an analysis of prognostic factors in five years survivors

Abstract

Aim: To evaluate the long-term survival benefit of TACE in patients with inoperable hepatocellular carcinoma and determine prognostic factors by analysis of actual 5-year survivors.

Methods: Pretreatment variables were analysed from a prospective database of 335 consecutive patients from January 1989 to December 1996. Univariate and multivariate analyses were performed to identify factors predictive of 5-year survival.

Results: Complete 5-year follow-up (median 91 months) was obtained on 320 patients (275 males) who underwent a mean of 3 TACEs (range 1–41) for inoperable HCC. Median age was 59 years (range 19–83). The majority were Child-Pugh grade A (81 per cent). Median tumour size was 9 cm (1–28 cm). There were 98 (30.6 per cent) 1-year and 36 (11.3 per cent) 3-year survivors. Among the 25 (7.8 per cent) 5-year survivors, survival ranged from 60.7 to 138.8 months (median 72.3 months) and tumour size from 1 to 21 cm (median 5.3 cm), with eight tumours greater than 10 cm. On univariate analysis, female gender (P = 0.037), absence of ascites (P = 0.028), platelet count <150 × 109 L−1 (P = 0.011), albumin >35 g L−1 (P = 0.04), α-fetoprotein <1000 ng mL−1 (P = 0.007), absence of venous invasion (P = 0.011), unilobar distribution (P = 0.027), less than three tumours (P = 0.015), and tumour size <8 cm (P = 0.021) were significant predictors of 5-year survival. Albumin >35 g L−1 (P = 0.014), unilobar distribution (P = 0.011) and α-fetoprotein <1000 ng mL−1 (0.014) were independent predictors of 5-year survival on multivariate regressional analysis.

Conclusion: Five-year survival is possible in patients with inoperable hepatocellular carcinoma. Poor prognosis is associated with a bilobar distribution, an α-fetoprotein level of >1000 ng L−1 on presentation and poor underlying liver function.