To evaluate the impact of a microscopically positive resection margin (microPRM) on the outcome of hepatoblastoma patients pretreated with chemotherapy.
Local recurrence and survival rates of 431 children treated in the SIOPEL 2 and 3 trials were analysed comparing 58 patients with microPRM with 371 who had a complete resection (CR) and who were then stratified by risk category. The tumour was standard-risk in 312 patients and high-risk (PRETEXT IV and/or extrahepatic disease and/or α-fetoprotein [AFP]<100 ng/ml) in 117 patients. All received cisplatinum-based neoadjuvant and postoperative chemotherapy as per protocol. Apart from one microPRM patient who went on to transplant, none received any additional local treatment.
With a median follow-up of 67 months, local relapse occurred in 3/58 patients with microPRM (5%) and in 23/371 (6%) patients with CR. The 5-year overall survival (OS) was 91% (95% confidence interval [CI] 80%-96%) for the microPRM and 92% (95% CI 89%-95%) for the CR group. The 5-year event-free survival (EFS) was 86% (95% CI 74%-93%) for the microPRM and 86% (95% CI 82%-89%) for the CR group. Neither OS nor EFS was statistically significantly different between the two groups, neither overall nor when risk group stratified.
In the context of cisplatin-based chemotherapy, the presence of microPRM did not influence the outcome even without additional local treatment. Although CR remains the aim, microPRM does not necessitate mandatory second look surgery. A 'wait and see policy' is warranted if postoperative chemotherapy is administered and AFP levels and imaging become normal.