Molecular evidence for a clonal relationship between multiple lesions in patients with unknown primary adenocarcinoma.

Unknown primary adenocarcinoma (UPA) comprises a group of heterogeneous cancers of great clinical and biological interest. UPA presents as metastatic disease without a detectable primary site after medical workup. Here we investigated whether or not a clonal relationship exists between multiple tumors within individual UPA patients. A molecular resemblance would argue for an early clonal outgrowth of tumor cells from the primary lesion, a mutual feature observed within this group of neoplasms. In 14 patients with UPA multiple tumors, obtained at autopsy, were analyzed by molecular allelotyping and immunohistochemistry. In addition, tumors of 4 patients could be analyzed by comparative genomic hybridization (CGH). Similar genetic and phenotypic profiles were used as indicator for a clonal relationship, whereas different profiles implicate independent tumors. The molecular data indicated that the multiple lesions in the 14 UPA patients, including the primary tumors, are clonally related. In agreement with the theory of tumor progression, some metastatic lesions showed additional genetic alterations besides the characteristics that were shared with the primary tumor. Furthermore, 8 UPA patients had tumors with a high frequency of allelic loss and/or imbalance (FALI; 43-71%), while 6 patients demonstrated a lower FALI (14-29%), suggesting the occurrence of chromosomal instability in the former group. Our data provide molecular evidence for a clonal relationship between multiple metastases and the primary tumor within individual UPA patients, independent of the anatomical origin of the cancer. This finding is in agreement with the suggestion that tumor progression is rapid in UPA patients, limiting the chance of clonal divergence. The identification of 2 groups of UPAs with either a high or low FALI indicates that chromosomal instability is not the only driving force behind early tumor cell dissemination. Thus, other molecular mechanisms must underlie the common biology of these tumors.