• National Cancer Institute
  • National Human Genome Research Institute

TCGA Tissue Sample Requirements: High Quality Requirements Yield High Quality Data

The Cancer Genome Atlas’s (TCGA) strict sample criteria are directly related to data quality.  These are some of TCGA’s rigorous sample requirements that yield high quality data. 

  • TCGA is a program of ‘no platform left behind’
    • TCGA works to ensure all research groups have equal access to TCGA data, and so has adopted the principle that no platform will be left behind.  This means that every sample must be analyzed through each platform.  Therefore, a sufficient sized sample is crucial to the TCGA pipeline.
    • For that reason, a resection sample is needed as opposed to a biopsy sample. Biopsies do not currently provide enough material (tissue) from which the RNA, and DNA are isolated to be used across all the different platforms.   As new approaches allow smaller inputs (less RNA and DNA, for example), the required tissue sample size requirements are expected to decrease.
  • Second-generation sequencing platforms facilitate use of more heterogeneous samples
    • Previously, tumor samples must have been composed of at least 80 percent tumor nuclei.  In other words, 80 percent of the cells in the sample must have been cancer cells.  However, with the power of second-generation sequencing platforms, TCGA is now able to investigate samples with only 60 percent tumor nuclei.  By decreasing the lower boundary, more samples are part of the TCGA program.  TCGA has found that 60 percent provides a sufficient proportion to generate high quality data in which the tumor’s signal can be distinguished from other cells’ signals.
    • TCGA is actively exploring how to utilize second-generation sequencing platforms to characterize highly heterogeneous tumors – such as pancreatic adenocarcinoma and diffuse gastric cancer.  These tumor types very rarely meet TCGA’s standard criteria because there are not sufficient cancer cells in each sample.  In an effort to characterize these highly heterogeneous cancers, TCGA is working to sequence more thoroughly to identify alterations found in very few cells.
  • Neoadjuvant treatment is not allowable
    • TCGA is interested in understanding the underlying genomics of primary, untreated tumors.  Cancer treatment can involve mutagens or carcinogens which could cloud the origin of the cancer.
  • Sample from primary tumor is necessary
    • A primary tumor is the tumor at the initial site of cancer, not where the cancer may have spread or metastasized, called the secondary tumor.  TCGA rarely studies this type of tumor, unless matched to a primary as part of a series (germline, primary and metastasis) or if the primary tumors are rarely diagnosed (such as in melanoma).
  • Tumor samples are paired with a source of germline DNA 
    • TCGA researchers need a second sample, discrete from the primary tumor, from the same patient.   Most often, this sample comes from blood to avoid any potential field effects in solid tumors.  The matched sample allows researchers to compare the tumor sample to the germline DNA. In other words, the blood or germline sample serves as a normal comparison of the tumor’s genome. 
  • Biospecimens are frozen
    • Samples must be frozen soon after surgery.  Samples need to be frozen quickly in order to prevent degradation of the RNA and DNA.  FFPE (formalin fixed paraffin embedded) samples are not suitable for molecular analysis because the RNA and DNA are trapped in the nucleic acid-protein cross linking from the fixation process.

These criteria are general guidelines of TCGA’s sample requirements in order to continue to produce high quality data. We welcome additional collaborators at any time as we recognize the need to involve teams from around the world to acquire the samples and data necessary for a program of this size and scope. For more information, email tcga@mail.nih.gov.