Learning to Negotiate the Genomic Complexities of Cancer

Get Permission

The data made available by TCGA represent fantastic opportunities for exploration and deeper understanding of the molecular drivers of cancer. It is now time to translate these into clinical significance.

—Joyce F. Liu, MD, MPH

The Cancer Genome Atlas (TCGA) Research Network recently reported the results of an integrated analysis of the genomic features of 373 endometrial carcinomas.1 This report joins previously published results of similar analyses in ovarian, breast, and colorectal cancers, squamous cell carcinoma of the lung, and glioblastoma as well as, most recently, acute myeloid leukemia.2-6

This ongoing project—a multicenter collaborative effort spearheaded by the National Cancer Institute and National Human Genome Research Institute—is planned to ultimately genomically characterize over 20 cancers, and all data generated from these efforts is publically available through a data portal at As the data from these analyses become available, they provide valuable insight into each of the target cancers, and will also serve as roadmaps as we learn to negotiate the genomic complexities of each of these diseases.

Validation and New Insights

The TCGA report on endometrial carcinoma both substantiates some of the clinical differences previously observed between endometrial cancer subtypes and provides new insights into their molecular nature. For example, serous carcinomas of the endometrium, which have been described to have a clinically different course from endometrioid endometrial cancers, indeed form a distinct molecular subtype with significantly worse outcomes, with the overwhelming majority (94%) of serous carcinomas demonstrating high copy number alterations, as well as mutations in TP53 (> 90%). Interestingly, about a quarter of high-grade (grade 3) endometrioid endometrial cancers also fall into this “serous-like” category, a molecular profile that shares similarities with that of high-grade ovarian carcinomas and triple-negative breast cancers.

TCGA also provides greater insight into the spectrum of type I endometrial cancers, separating the non–­serous-like tumors into three additional clusters—an ultramutated group with POLE mutations, a hypermethylated group with microsatellite instability, and a copy number–low group. Additionally, the analysis reinforces some of the previously described molecular findings in endometrial cancers, such as a high rate of PI3K/AKT pathway alterations, including mutations in PTEN and PIK3CA, as well as mutations in genes such as ARID1A and KRAS, and also describes novel mutations in genes such as ARID5B.

Frequent activation of the WNT/CTNNB1 signaling pathway was also observed, with a characteristic exclusivity between KRAS and CTNNB1 mutations suggesting that alternative mechanisms of WNT signaling activation are present in endometrioid endometrial cancers compared to other tumors, such as colon cancer.

Questions and Challenges

The findings from this study provide us with a wealth of information about the molecular spectrum of endometrial cancer. As with other TCGA studies, the real questions and challenges as we move forward will lie in how we apply this knowledge. These data suggest many promising pathways for exploration, including potential novel therapies and new approaches to therapy. For example, should we identify the subset of “serous-like” endometrioid cancers and recommend adjuvant chemotherapy to these patients as we do for serous cancers? Can we improve clinical outcomes by treating patients with specific pathway alterations with drugs that target the relevant pathway?

Although TCGA provides molecular evidence to support such efforts, the link between molecular characterization and clinical outcome is not yet clear. For example, attempts to link targeted therapy responses to pathway alterations have not been easily interpreted in endometrial cancer. Janku et al reported on a cohort of 140 patients with breast, cervical, endometrial, or ovarian cancer referred to the Clinical Center for Targeted Therapy at The University of Texas MD Anderson Cancer Center.7 Patients on study were analyzed for PIK3CA, KRAS, NRAS, and BRAF mutations; 25 were found to have PIK3CA mutations and, of these, 23 were treated on a protocol that included a PI3K/AKT/mTOR pathway inhibitor. A 30% partial response rate was observed in these patients, compared to 10% in 70 patients with wild-type PIK3CA treated on the same protocols (P = .04). In contrast, in a phase II study of the AKT-1/2/3 inhibitor MK-2206 in endometrial cancer, Myers et al reported no clear correlation between PIK3CA mutation and clinical response.8

Translation to Clinical Practice

These results highlight the challenges that lie ahead as we seek to integrate the growing stream of genomic and molecular information into clinical practice. Even a potential biomarker that seems relatively straightforward, such as PIK3CA mutation, may not prove to be so in the setting of multiple interacting pathways and escape mechanisms.

The data made available by TCGA represent fantastic opportunities for exploration and deeper understanding of the molecular drivers of cancer. It is now time to translate these into clinical significance. As we analyze the data from TCGA analysis of endometrial cancer, it is therefore even more imperative that we incorporate molecular analyses into our clinical trials, to better determine how these molecular subtypes should influence our treatment strategies. ■

Dr. Liu is Instructor in Medicine at Harvard Medical School and Dana-Farber Cancer Institute, Boston. 

Disclosure: Dr. Liu reported no potential conflicts of interest.


1. Cancer Genome Atlas Research Network: Integrated genomic characterization of endometrial carcinoma. Nature 497:67-73, 2013.

2. Cancer Genome Atlas Research Network: Integrated genomic analyses of ovarian carcinoma. Nature 474:609-615, 2011.

3. Cancer Genome Atlas Network: Comprehensive molecular portraits of human breast tumours. Nature 490:61-70, 2012.

4. Cancer Genome Atlas Research Network: Comprehensive genomic characterization of squamous cell lung cancers. Nature 489:519-525, 2012.

5. Cancer Genome Atlas Research Network: Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455:1061-1068, 2008.

6. Cancer Genome Atlas Research Network: Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med 368:2059-2074, 2013.

7. Janku F, Wheler JJ, Westin SN, et al: PI3K/AKT/mTOR inhibitors in patients with breast and gynecologic malignancies harboring PIK3CA mutations. J Clin Oncol 30:777-782, 2012.

8. Myers AP, Broaddus R, Makker V, et al: Phase II, two-stage, two-arm, PIK3CA mutation stratified trial of MK-2206 in recurrent endometrial cancer (EC). J Clin Oncol 31(suppl):Abstract 5524, 2013.

Related Articles

ntegrated Genomic Characterization of Endometrial Carcinomas Suggests New Classification Scheme

As recently reported in Nature, investigators in The Cancer Genome Atlas Research Network performed an integrated genomic, transcriptomic, and proteomic characterization of 373 endometrial carcinomas, including 307 endometrioid and 66 serous or mixed histology cases, using array- and...