Clonality analysis of cervical cancer on microdissected archival materials by PCR-based X-chromosome inactivation approach.

ABSTRACT

Clonality of invasive human cervical cancer was assessed in microdissected archival formaldehyde-fixed, paraffin-embedded tissues by PCR-based analysis of X-chromosome inactivation of the androgen receptor gene. In 18 informative cases, cancers from 12 cases including 2 early-invasive cancers were scored as monoclonal. Among them, 8 cases had the combination of random X-chromosome inactivation in normal cervical tissue and non-random in the cancer and in other 4 cases normal and cancer tissue were both non-random but had discordant X-chromosome inactivation. Six cases showed that the non-random inactivation affected the same allele in cancer and normal tissue and thus were considered to be inconclusive. The results suggest the monoclonal origin of cervical cancer and indicate that genetic events are critical in transition of pre-malignant epithelia to invasive cancer in cervical carcinogenesis. Finding of monoclonal early invasive cancer also argues that monoclonality of human tumors is not a late event due to clonal competition or selection. X-chromosome inactivation patterns in normal cervical epithelia and tissues were analyzed in 21 informative cases. When a mixture of normal stroma and epithelium was analyzed, 38% (8/21) of cases showed non-random inactivation pattern, whereas using pure epithelium it was 64% (11/17). The X-chromosome inactivation pattern between mixed epithelium/stroma and matched epithelium differed in 2/8 cases. These results point to a certain amount of overlooked source of error in X-inactivation-based tumor clonality analysis in which peripheral blood monocytes were chosen as control and strongly recommend that control tissues which share close histological origin with analyzed tumors should be selected in any clonality study of human neoplasm. Post-embryological mechanisms of clonal patch in normal cervical epithelia are also discussed.