A comprehensive analysis of endometrial cancer from The Cancer Genome Atlas Research Network suggests that tumors can be divided into 4 distinct subtypes based on their molecular profiles, potentially leading to more accurate diagnosis and better-informed treatment choices.

The most comprehensive analysis of endometrial tumors to date suggests that they can be divided into four distinct subtypes based on their genomic profiles, potentially leading to a more accurate diagnosis of the disease and better choice of therapies, including targeted drugs currently being tested in clinical trials.

Conducted by investigators within The Cancer Genome Atlas Research Network, the analysis also identified genomic similarities between endometrial cancer and breast, ovarian, and colorectal cancers (Nature 2013; 497:67–73).

Clinically, endometrial cancers fall into 2 groups: endometrioid (type I) and serous (type II) tumors. Type I tumors are linked to excess estrogen, obesity, and a favorable prognosis. Type II tumors are more common in older women and tend to have a worse outcome. In addition to surgery, patients with early-stage endometrioid tumors typically undergo radiation therapy, while those with serous and late-stage endometrioid tumors typically receive chemotherapy.

However, “higher-grade endometrioid and serous tumors are tough to distinguish from each other,” says Douglas Levine, MD, head of the Gynecologic Research Laboratory at Memorial Sloan-Kettering Cancer Center and coleader of the study. “About 25% of the time, pathologists can't agree on what to call these tumors,” he says, and that difference of opinion can affect treatment choice.

In the study, researchers performed an integrated genomic and proteomic analysis of tumor samples from 373 patients. About 25% of the tumors pathologists classified as high-grade endometrioid had frequent mutations in TP53 and numerous copy number alterations, characteristics associated with serous tumors. In contrast, most other endometrioid tumors had few TP53 mutations or copy number alterations, but they did have frequent mutations in other cancer-associated genes, including PTEN, CTNNB1, PIK3CA, ARID1A, and KRAS.

Based on their results, the researchers divided the endometrial tumors into four groups:

  • POLE ultramutated, which exhibited high mutation rates in the POLE gene and accounted for about 10% of endometrioid tumors

  • hypermutated microsatellite instability, which had high rates of mutations but no POLE mutations

  • copy-number low, which had a high frequency of mutations in CTNNB1 and increased progesterone receptor activity, suggesting responsiveness to hormone therapy

  • copy-number high, composed mostly of serous tumors, almost all of which had TP53 mutations. Some high-grade endometrioid tumors fall into this group, indicating that they may need to be treated more aggressively after surgery.

In addition, the researchers found that high-grade uterine serous tumors share many molecular features with high-grade serous ovarian tumors and basal-like breast carcinomas, hinting that they could be treated similarly. Likewise, endometrioid endometrial carcinomas share some characteristics of colorectal cancers, including microsatellite instability and POLE mutations.

The findings could have clinical implications for patients within the next year or two, says Levine. For example, patients with POLE mutations had excellent outcomes in the study, so they may not need additional treatment following surgery. However, this group included only 17 patients, so further study is needed to confirm the finding.

Levine and others are now enrolling patients with endometrial cancer in clinical trials of targeted therapies based on their cancer's molecular traits, paralleling earlier work in breast cancer, with more trials to follow. “We need to confirm that certain subtypes respond better to particular targeted therapies so we can improve outcomes for these women,” he says.

For more news on cancer research, visit Cancer Discovery online at http://CDnews.aacrjournals.org.