github link
Accession IconGSE60542

Revisiting the transcriptional analysis of primary tumors and associated nodal metastases with enhanced biological and statistical controls: application to thyroid cancer

Organism Icon Homo sapiens
Sample Icon 88 Downloadable Samples
Technology Badge Icon Affymetrix Human Genome U133 Plus 2.0 Array (hgu133plus2)

Submitter Supplied Information

The biology underlying nodal metastasis is poorly understood. Transcriptome profiling has helped to characterize both primary tumors seeding nodal metastasis and the metastasis themselves. The interpretation of these data, however, is not without ambiguities. Here we profiled the transcriptomes of 17 papillary thyroid cancer (PTC) nodal metastases, associated primary tumors and primary tumors from N0 patients. We also included patient-matched normal thyroid and lymph node samples as controls to address some limits of previous studies. We found that the transcriptomes of patient-matched primary tumors and metastases were more similar than of unrelated metastases/primary pairs, a result also reported in other organ systems, and that part of this similarity reflected patient background. We found that the comparison of patient-matched primary tumors and metastases was heavily confounded by the presence of lymphoid tissues in the metastasis samples. An original data adjustment procedure was developed to circumvent this problem. It revealed a differential expression of stroma-related gene expression signatures also regulated in other organ systems. The comparison of N0 vs. N+ primary tumors uncovered a signal irreproducible across independent PTC datasets. This signal was also detectable when comparing the normal thyroid tissues adjacent to N0 and N+ tumors, suggesting a cohort specific bias also likely to be present in previous studies with similar statistical power. Classification of N0 vs. N+ yielded an accuracy of 63%, but additional statistical controls not presented in previous studies, revealed that this is likely to occur by chance alone. To address this issue, we used large datasets from The Cancer Genome Atlas and showed that N0 vs. N+ classification rates could not be reached randomly for most cancers. Yet, it was significant, but of limited accuracy (<70%) for thyroid, breast and head and neck cancers.
PubMed ID
Total Samples
Submitter’s Institution
Alternate Accession IDs


Show of 0 Total Samples
Accession Code
Processing Information
Additional Metadata
No rows found