168P - Response assessments in hepatocellular carcinoma: What are the best criteria to utilize? mRECIST or RECIST 1.1? A retrospective meta-analysis of multiple phase III trials

Background

In 2010 mRECIST was published in Seminars of Liver Disease for the evaluation of hepatocellular carcinoma (HCC) response assessment. It proposed a modification of existing RECIST 1.1 incorporating visible changes and reduction in the arterial enhancement of the HCC tissue and the concept of measuring viable versus not measuring necrotic tissue, better evaluating treatment responses in HCC with CT or MRI.

Methods

Multiple phase III clinical trials were analyzed using both mRECIST and RECIST 1.1 in parallel, which were read separately by blinded independent central read (BICR). The purpose was to compare the overall response rates and complete response rates by the two criteria. 2601 subjects with 13675 post-baseline imaging timepoints were included in this comparison. The Overall Response Rate (ORR) and the Complete Response Rate (CRR) are derived.

Results

Table: 168P

Overall response in 2601 subjects with 13675 time points excluding baseline

Conclusions

Our results indicate that: 1.) mRECIST has better ORR and CRR compared to RECIST 1.1. 2.) RECIST 1.1 has more timepoints with Stable Disease SD/NN compared to mRECIST. 3.) mRECIST has more Not Evaluable (NE) timepoints due to more stringent imaging specifications. mRECIST shows almost double the CRR than RECIST 1.1, and the ORR is 33% higher using mRECIST than RECIST 1.1. Stable Disease was observed about 20% more in the RECIST 1.1 analysis, which includes the Non-CR/Non-PD assessment of patients with only non measurable disease for non-target lesions. A reasonable explanation for such a 20% SD discordance being higher with RECIST 1.1 is primarily reflected in the higher number of documented responders assessed with mRECIST. A Non Evaluable (NE) response is more common with mRECIST. Our analysis confirms that mRECIST is more robust in assessing responding patients with CRR and ORR in HCC as it based on measuring the viable tumor component instead of plain linear RECIST 1.1 measurements of its total tumor diameter only.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.