Circulating biomarkers in hepatocellular carcinoma
Purpose Our aims are to determine levels of circulating cellular and protein biomarkers in hepatocellular carcinoma (HCC) patients and to analyse any relationships with clinical parameters.
Methods Fifty-four consenting patients were recruited.Circulating tumour cells (CTCs) were enumerated (by Cell-Search) and characterised via filtration [by isolation by size of epithelial tumour cells (ISET)] with downstream immunohistochemistry (IHC). Glypican-3 (GPC3) expression in tumour biopsies and CTCs (by IHC) was compared, and levels of circulating caspase-cleaved and full-length cytokeratin 18 (CK18, measured using M30 and M65 ELISAs) were examined as a putative prognostic factor and marker of tumour burden.
Results CTCs were identified in 14 out of 50 (28 %) patients by CellSearch and in 19 out of 19 (100 %) patients by ISET. The presence of GPC3-positive CTCs by ISET was 100 % concordant with the presence of GPC3-positive cells in the original tumour (n = 5). No statistically significant correlations were observed between CTC number and clinical characteristics, although trends were noted between CTC subtypes, Child–Pugh score and tumour node metastasis stage. Serum M30 and M65 levels (as continuous variables) significantly correlated with overall survival (OS)in a univariate analysis (p = 0.003 and p < 0.001, respectively);M65 levels remained statistically significant in a multivariate analysis (p = 0.029).
Conclusions This is the first study to detect GPC3-positive CTCs in HCC, important for drug development with this target. The significant association of circulating CK18 with OS in HCC further exemplifies the utility of circulating biomarkers in cancer.
Materials and methods
Sequential patients with a diagnosis of HCC were recruited to the study at two centres: The Christie NHS Foundation Trust, Manchester, UK, and North Manchester General Hospital, Manchester, UK. Patients either had no antitumour therapy within 4 weeks, or had evidence of progressive disease if prior therapy had been received and were awaiting further therapy. Clinical data collected included age, sex, Child–Pugh score, serum alpha-fetoprotein (AFP), performance status (PST), viral status, radiological extent of disease, time to progression and overall survival (OS). In addition, treatment history, tumour node metastasis (TNM) staging, Okuda stage, Barcelona stage (BCLC) and aetiological risk factors were collated.
After informed consent, a single blood draw of up to 30 ml was taken for biomarker analysis at baseline (defined as 4 weeks after any previous treatment, or evidence of progressive disease if therapy had been received and awaiting further therapy or supportive care only). Serum was isolated from 10 ml coagulated whole blood (collected in serum gel tubes) for analysis of the levels of full-length and/or caspase-3-cleaved cytokeratin 18 (CK18) using the M65 and M30 ELISAs, respectively; 10 ml blood was sampled into a CellSave tube (Veridex) for CTC enumeration by CellSearch, and wherever possible 10 ml (EDTA ) was sampled for analysis using the filtration-based ISET device. Tumour biopsies were also obtained from 12 patients in order to compare GPC3 expression levels in CTCs with tumour at the time of diagnosis.
Cell lines and antibodies
SK-HEP-1 HCC cells (GPC3-negative) that had been transfected with human GPC3 cDNA were used as positive controls:SK-03 was a high expresser of GPC3, and SK-PCA13a had intermediate expression of GPC3 . These cell lines were provided by Chugai Pharmaceutical Co., Ltd. Human Vascular Endothelial Cells (HUVEC) were used as endothelial cell controls to allow exclusion of circulating endothelial cells (CECs) in the CTC assay following CTC enrichment by ISET. All cell lines were cultured according to standard protocols. A mouse monoclonal antibody against GPC3 (mGC33) for immunohistochemical (IHC) analysis of tissue biopsies and of CTCs captured on ISET filters was provided by Chugai Pharmaceutical Co., Ltd. . The cell lines and mGC33 antibody were used to validate GPC3 immunohistochemical assays in tissue and in cells.
Enumeration and assessment of GPC3 status of CTCs
CTCs were enumerated using the CellSearch platform in blood samples from 50 patients. The CellSearch methodology has been described previously [11, 12] and is based on the immunomagnetic capture of epithelial adhesion molecule (EpCAM) positive cells followed by immunophenotyping of cytokeratin, DAPI and CD45 in cells isolated from a 7.5-ml blood volume. CTCs are classed as CK+,DAPI+ and CD45?. In order to establish GPC3 expression in CTCs, an AF-488-conjugated antibody against GPC3 (mGC33) was also added to the samples in the fourth channel of the CellSearch at a concentration of 4 μg/test. The exposure time for signal detection of GPC3 was 0.4 s. This enabled the enumeration of GPC3-positive CTCs as a proportion of the total CTC count in 7.5 ml blood.