Endometrial Carcinoma: Molecular Subtypes and Treatment

Endometrial carcinoma is cancer that begins in the lining of the uterus and accounts for the majority of uterine cancers diagnosed worldwide. It is often caught early because its hallmark symptom, abnormal vaginal bleeding, tends to appear before the disease has spread. But the term “endometrial carcinoma” covers a surprisingly wide range of tumors, from slow-growing cancers driven by excess estrogen to aggressive subtypes that behave more like ovarian or fallopian tube cancers. Advances in molecular profiling over the past decade have reshaped how the disease is classified, staged, and treated.

How the Disease Has Been Traditionally Classified

For decades, endometrial cancers were split into two broad buckets. Type I tumors, the more common group, are estrogen-dependent, usually low-grade endometrioid cancers that grow slowly and carry a relatively favorable prognosis. Type II tumors are estrogen-independent, clinically aggressive, and include serous carcinomas, clear cell carcinomas, and other high-grade variants.¹ This binary framework was useful as a rough guide, but it papered over a lot of biological diversity. Two tumors that look identical under the microscope can carry very different mutations and behave very differently in patients. That gap between appearance and biology is what molecular classification now addresses.

Four Molecular Subtypes

The Cancer Genome Atlas project analyzed hundreds of endometrial tumors and identified four distinct molecular groups, each with its own mutation burden, genomic features, and clinical trajectory.² These subtypes have since become central to how oncologists think about prognosis and treatment selection.³

POLE ultramutated: These tumors carry mutations in a DNA repair gene called POLE and have extraordinarily high mutation rates. Despite sounding alarming, they tend to have the best outcomes, likely because the heavy mutation load triggers a strong immune response against the cancer.
Microsatellite instability hypermutated (MSI-H): These cancers have defects in mismatch repair, a separate DNA maintenance system. Most arise from silencing of the MLH1 gene. They also carry high mutation burdens, though lower than the POLE group, and often respond well to immunotherapy.
Copy number low: The largest group, comprising most microsatellite-stable endometrioid tumors. These have a moderate mutation rate and an intermediate prognosis.
Copy number high: Dominated by serous-like cancers with extensive chromosomal rearrangements. This group has the worst prognosis and overlaps substantially with what was traditionally called Type II disease.

The updated 2023 FIGO staging system for endometrial cancer formally incorporated molecular classification, meaning a tumor’s molecular subtype can now influence its official stage assignment.⁴ A POLE-mutated tumor that looks high-grade under the microscope, for instance, may be downstaged because of its favorable biology. Implementing this in practice has posed challenges, particularly around standardizing POLE mutation testing across institutions.⁵

Risk Factors and Who Gets It

The dominant risk factor for endometrial carcinoma is prolonged, unopposed estrogen exposure to the uterine lining. Obesity is the single biggest modifiable contributor because fat tissue produces estrogen. In postmenopausal women, this extra estrogen acts directly on the endometrium. In premenopausal women, the mechanism is slightly different: obesity promotes chronic high insulin levels and excess ovarian androgen production, which can suppress ovulation and leave the endometrium exposed to estrogen without the counterbalancing effect of progesterone.⁶

Polycystic ovary syndrome (PCOS) raises risk through a similar pathway. Women with PCOS frequently have anovulatory cycles, meaning they do not ovulate regularly and therefore produce little progesterone. Without progesterone to counteract estrogen-driven growth, the endometrium can develop hyperplasia and eventually cancer over time.⁷

Other factors that increase exposure to estrogen, like taking estrogen-only hormone replacement therapy without progesterone, early onset of menstruation, late menopause, and tamoxifen use for breast cancer, also elevate risk. Diabetes, metabolic syndrome, and a sedentary lifestyle compound the picture, often alongside obesity.

Lynch Syndrome and Hereditary Risk

Roughly three to five percent of endometrial cancers are linked to Lynch syndrome, an inherited condition caused by mutations in mismatch repair genes such as MSH2, MLH1, MSH6, and PMS2.⁸ Women with Lynch syndrome face a lifetime endometrial cancer risk of roughly 40 to 60 percent, which equals or exceeds their risk of colorectal cancer. For women carrying MSH6 mutations specifically, the cumulative risk climbs even higher, reaching about 71 percent by age 70.⁹

Endometrial cancer may be the first cancer diagnosed in women with Lynch syndrome, sometimes appearing years before any colon cancer. That’s why universal screening of newly diagnosed endometrial tumors for mismatch repair deficiency has become standard at many cancer centers. A finding of deficient mismatch repair not only triggers genetic testing and cancer surveillance for the patient and her family, but also flags the tumor as potentially responsive to immunotherapy.

Oral Contraceptives as a Protective Factor

On the prevention side, combined oral contraceptives offer meaningful long-term protection. A large individual-level meta-analysis covering more than 27,000 women with endometrial cancer found that every five years of oral contraceptive use reduced risk by about 24 percent, and that protection persisted for more than 30 years after stopping the pill.¹⁰ The protective effect held regardless of whether the pills were the higher-dose formulations used in earlier decades or the lower-dose versions available later. The mechanism is straightforward: combined pills contain progestin, which counteracts estrogen’s stimulating effect on the endometrial lining.

How It Is Detected and Diagnosed

There is no routine screening test for endometrial cancer in the general population, which distinguishes it from cancers like cervical or breast cancer. The most common presenting symptom is abnormal uterine bleeding, particularly any vaginal bleeding after menopause. When a woman with postmenopausal bleeding undergoes transvaginal ultrasound, an endometrial thickness below about 4 to 5 millimeters makes significant endometrial disease unlikely. A large Nordic study established that when the endometrium measured 4 mm or less, the risk of finding abnormal tissue at curettage was only about 5.5 percent.¹¹ A systematic review and meta-analysis later found that using a 5 mm cutoff maintained high sensitivity for detecting cancer while reducing unnecessary invasive procedures by about 17 percent compared with a 3 or 4 mm threshold.¹²

When the endometrium is thick or bleeding persists, tissue sampling is needed. The standard approach is an office endometrial biopsy, often performed blindly with a pipelle device. But biopsy under direct visualization through hysteroscopy yields better sample adequacy and is less likely to miss disease. A meta-analysis found that hysteroscopy-guided biopsy had significantly higher rates of adequate samples and a much lower risk of failing to detect cancer or hyperplasia compared with blind sampling.¹³ In practice, blind office biopsy remains the first step at most centers because it is quick, inexpensive, and reasonably accurate. Hysteroscopy is typically reserved for cases where blind biopsy is inconclusive or insufficient.

Surgery and Sentinel Lymph Nodes

Surgery is the foundation of treatment for endometrial carcinoma. The standard operation is a hysterectomy with removal of both fallopian tubes and ovaries, performed minimally invasively (laparoscopy or robotic-assisted surgery) whenever possible. One of the biggest shifts in surgical management over the past decade has been the adoption of sentinel lymph node mapping. Instead of removing large chains of pelvic and para-aortic lymph nodes, surgeons inject a tracer dye into the cervix to identify the first nodes draining the tumor, then remove and carefully examine those sentinel nodes.

Evidence shows that sentinel lymph node mapping in early-stage, lower-risk endometrial cancer has a high detection rate, strong sensitivity, and an excellent negative predictive value, meaning that if the sentinel nodes are cancer-free, it is very unlikely that other nodes harbor disease. Importantly, patients who underwent sentinel node mapping rather than full lymphadenectomy did not have worse progression-free or overall survival, and they experienced fewer surgical complications.¹⁴ Fewer removed lymph nodes also means a lower risk of lymphedema, a chronic swelling condition that significantly worsens quality of life in endometrial cancer survivors.¹⁵

Radiation After Surgery

Many women with early-stage endometrial cancer at intermediate or high-intermediate risk receive adjuvant radiation to reduce the chance of local recurrence. For years, the default was external beam radiation therapy (EBRT) targeting the entire pelvis. But the PORTEC-2 trial demonstrated that vaginal brachytherapy, a much more targeted treatment where a radiation source is placed inside the vaginal canal for brief sessions, achieved equivalent vaginal recurrence control. Five-year vaginal recurrence rates were under two percent in both groups. The decisive difference was in side effects: gastrointestinal toxicity at the end of treatment was dramatically lower with brachytherapy, affecting about 13 percent of patients compared with 54 percent receiving external beam therapy.¹⁶ Vaginal brachytherapy is now considered the standard adjuvant treatment for this risk group.

Immunotherapy for Mismatch Repair-Deficient Tumors

The discovery that a substantial fraction of endometrial cancers are mismatch repair-deficient opened the door to checkpoint immunotherapy. Dostarlimab, an antibody that blocks the PD-1 protein on immune cells, has shown strong activity in recurrent or advanced endometrial cancer with mismatch repair deficiency. In a phase 1 trial, the response rate was about 42 percent, with roughly 13 percent of patients achieving a complete response. Responses were durable: the estimated likelihood of maintaining a response at 12 months was about 77 percent.¹⁷ A real-world multicenter study confirmed these findings, reporting an objective response rate of about 54 percent and complete responses in roughly 22 percent of patients, with the majority maintaining their response at both 12 and 24 months.¹⁸

Pembrolizumab, another PD-1 inhibitor, has been approved in combination with the targeted agent lenvatinib for advanced endometrial cancer regardless of mismatch repair status, and as a single agent for mismatch repair-deficient tumors. The combination has broadened immunotherapy’s reach to the copy number low and copy number high groups that don’t respond as well to single-agent checkpoint blockade. These approvals have fundamentally changed how recurrent endometrial cancer is managed, especially for patients whose tumors progress after chemotherapy.

Targeted Therapy for HER2-Positive Serous Carcinoma

Uterine serous carcinoma, the most aggressive common subtype, overexpresses the HER2 protein in roughly a quarter to a third of cases. A randomized trial tested adding trastuzumab, an antibody that targets HER2, to standard carboplatin-paclitaxel chemotherapy in women with advanced or recurrent HER2-positive serous carcinoma. Median progression-free survival jumped from about 8 months to roughly 13 months with the addition of trastuzumab, and overall survival improved from about 24 months to nearly 30 months.¹⁹ A larger retrospective study reinforced these findings, showing that trastuzumab added to chemotherapy roughly doubled survival in advanced-stage disease.²⁰ Newer antibody-drug conjugates targeting HER2, Trop-2, and folate receptor alpha are also being tested in endometrial cancer and have shown encouraging early response rates.²¹

Hormonal Therapy in Recurrent Disease

For women with recurrent, low-grade endometrial cancer that still expresses estrogen and progesterone receptors, hormonal therapy remains an important option. Progestins, aromatase inhibitors, and other hormonal agents can achieve good disease control with far fewer side effects than chemotherapy. Because most endometrial cancers are hormone receptor-positive, this approach has broad applicability, though it works best in low-grade, hormone-driven tumors.²² Hormonal treatment can also be offered to patients who are not surgical candidates, or as a bridge in settings where more aggressive therapy is not tolerated.²³

Fertility Preservation for Younger Patients

Endometrial cancer is increasingly diagnosed in women of reproductive age, partly driven by rising obesity rates. Standard treatment involves hysterectomy, which obviously eliminates the possibility of pregnancy. For selected young women with early-stage, well-differentiated tumors confined to the endometrium without deep invasion, fertility-sparing hormonal treatment with oral progestins or a levonorgestrel-releasing intrauterine device is an established option.²⁴

Complete response rates with progestin therapy are encouraging, with guidelines citing figures in the range of 67 to 80 percent.²⁵ However, recurrence after initial response is common. In one Korean study following 148 women, about 78 percent achieved complete response to progestin treatment, but roughly 30 percent of those responders experienced recurrence during a median follow-up of about five and a half years.²⁶ For this reason, women who choose fertility preservation are counseled to pursue pregnancy promptly after achieving a response and to undergo hysterectomy once childbearing is complete.

Racial Disparities in Outcomes

Endometrial cancer survival differs starkly by race. Black women have significantly worse outcomes compared with white women, a gap that has persisted for decades. Analysis of national surveillance data found that these disparities are driven largely by a higher proportion of advanced-stage diagnoses, aggressive histologic subtypes, and differences in the use of surgery among Black patients.²⁷ The same analysis suggested that when Black women received more intensive surgical and radiation treatment, their survival improved, raising difficult questions about whether access to and receipt of appropriate treatment contribute to the gap alongside biologic differences in tumor subtype distribution.

AI and the Future of Molecular Subtyping

Molecular subtyping requires specialized testing, including immunohistochemistry for mismatch repair proteins, p53 sequencing, and POLE mutation analysis. Not every pathology lab has these capabilities, which creates inequities in access to precision staging. Researchers are exploring whether artificial intelligence can bridge this gap by predicting molecular subtypes directly from standard tissue slides stained with routine dyes.

A multicenter Chinese study developed a deep-learning system that classified all four molecular subtypes from routine slide images with accuracies around 90 to 92 percent and correlated well with observed patient survival.²⁸ Another model integrated MRI-based imaging features with clinical data to classify molecular subtypes noninvasively, before surgery even occurs.²⁹ Separate work using a multi-resolution neural network showed the ability to predict not just subtypes but individual gene mutations from digitized pathology images.³⁰ None of these tools are ready to replace standard molecular testing, but they could eventually serve as screening tools in settings where full genomic profiling is unavailable or as a rapid first pass to prioritize confirmatory testing.

The Microbiome Connection

An emerging and still-speculative area of research involves the role of the microbiome in endometrial cancer. The uterine cavity was long assumed to be sterile, but newer sequencing techniques have identified bacterial communities in the endometrium. Research suggests that microbial imbalance may contribute to endometrial cancer development through chronic inflammation and alterations in estrogen metabolism, and may also influence immune behavior within the tumor environment.³¹ The idea is still in early days, with most data coming from observational studies that can identify associations but not prove causation. Whether manipulating the microbiome could ever help prevent or treat endometrial cancer remains a genuinely open question. It is a field worth watching but not one ready to influence clinical decisions.