F. Negro et al., DETECTION OF HUMAN ANDROGEN RECEPTOR MESSENGER-RNA IN HEPATOCELLULAR-CARCINOMA BY IN-SITU HYBRIDIZATION, Liver, 14(4), 1994, pp. 213-219
Recent evidence suggests that anti-androgen therapy may be useful in p
atients with androgen receptor (AR)-positive hepatocellular carcinomas
(HCC), as determined by a steroid binding assay. To evaluate the AR e
xpression of HCC, in both histological and cytological material, we de
veloped a non-radioisotopic in situ hybridisation (NISH) assay specifi
c for the human AR mRNA. A synthetic oligonucleotide complementary to
positions 661-695 of the human AR coding sequence was end-labelled wit
h digoxigenin-dUTP and revealed by an alkaline phosphatase-conjugated
anti-digoxigenin antibody. We analysed 22 formalin-fixed, paraffin-emb
edded HCC, obtained at surgery, together with the corresponding non-ne
oplastic liver tissues (19 cases). In six cases, cell blocks obtained
by fine-needle aspiration (FNA) prior to surgery were also available.
Positive controls included seminal vesicles and prostate tissues. Sixt
een HCCs (73%) expressed a variable amount of AR mRNA, with the propor
tion of positive cells ranging from very few to more than 90%. Normal
hepatocytes were stained weakly and focally in eight cases (42%). Appr
opriate controls, inclusive of immunohistochemical detection of the AR
protein in selected cases, established the specificity of the assay.
Data obtained on FNA specimens were predictive of the results on histo
logic material. However, in two cases the NISH assay was negative on t
he cytological specimen but stained rare hepatocytes within the surgic
ally resected tumor. In conclusion, NISH is a novel procedure for rapi
d and specific assessment of the expression of AR in HCC tissue. Its c
linical significance, in terms of predictivity of response to anti-and
rogen treatment, needs to be assessed in large correlative studies.