Endometrial Cancer Staging: FIGO System and Molecular Types

Endometrial cancer staging is the process of determining how far the disease has spread beyond the uterine lining, and it drives virtually every treatment decision that follows. The staging system was overhauled in 2023 by the International Federation of Gynecology and Obstetrics (FIGO), replacing a framework that had been in use since 2009. The new system does something no previous version attempted: it folds molecular tumor profiling into the stage assignment itself, meaning two tumors that look identical under a microscope can now land in different stages based on their genetic characteristics.

How Staging Changed in 2023

The 2009 FIGO system relied almost entirely on anatomy. How deep did the tumor grow into the uterine wall? Did it reach the cervix? Were lymph nodes involved? Did it spread to distant organs? Those questions still matter, but the 2023 update adds layers of biological detail that the older system ignored. The revised framework integrates histology, tumor grade, lymphovascular space invasion, and molecular subtype to create substages with sharper prognostic accuracy.¹ The goal is straightforward: two patients at the “same” anatomical stage can have wildly different outcomes, and the old system couldn’t tell them apart. The new one tries to.

The previous system’s simplicity was both its strength and its weakness. A clinician anywhere in the world could assign a stage with a pathology report and surgical findings. But research accumulated over the 14-year gap showed that anatomy alone left too much prognostic information on the table.² The update attempts to fix that, though not without trade-offs.

Molecular Classification and What It Means for Your Stage

The most talked-about change is the incorporation of four molecular subtypes into the staging system. Tumors can now be classified as POLE-mutated, mismatch repair deficient (MMRd), no specific molecular profile (NSMP), or p53-abnormal. Each subtype carries a different prognosis. POLE-mutated tumors tend to behave well even when they look aggressive under a microscope, while p53-abnormal tumors tend to behave badly even when they appear early-stage.

When molecular testing is performed, the stage is annotated with an “m” plus a subscript indicating the subtype. And this isn’t just bookkeeping. In stages I and II, knowing the molecular subtype can actually change the stage assignment. A POLE-mutated tumor can be downstaged, while a p53-abnormal tumor can be upstaged.³ That shift has real consequences for whether someone gets additional treatment after surgery.

Patients with FIGO stage I non-p53-mutated tumors, and those with stage I or II POLE-mutated tumors, generally have excellent outcomes, and adjuvant therapy is often not recommended for them at all.⁴ On the other end, a p53-abnormal tumor that might have been called stage I under the old system could now be classified as stage II, triggering more aggressive treatment. The molecular layer essentially overrides what the tumor looks like physically when the biology tells a different story.

The Surgical Staging Procedure

Endometrial cancer is primarily staged through surgery, not imaging. The standard surgical staging procedure involves removal of the uterus, cervix, and both ovaries and fallopian tubes, along with sampling of pelvic and para-aortic lymph node tissue and collection of pelvic washings.⁵ Everything removed goes to the pathology lab, where the tumor’s depth of invasion, grade, histological type, and other features are assessed under the microscope. This pathology report is the foundation of the final stage.

The reason surgery is the gold standard is that no imaging technology can match the detail of a direct tissue examination. Preoperative scans give an educated guess; the surgical specimen gives the answer. That said, the extent of surgery itself is guided by what imaging and biopsy suggest beforehand, which creates a kind of feedback loop: you need to know roughly what you’re dealing with to plan the right operation, but you won’t know exactly what you’re dealing with until the operation is done.

Sentinel Lymph Node Mapping

One of the bigger shifts in surgical staging practice has been the adoption of sentinel lymph node mapping. Rather than removing large numbers of lymph nodes from the pelvis and around the aorta, surgeons inject a tracer dye near the tumor and identify the first nodes that drain the area. If those sentinel nodes are clean, the odds of cancer having spread to other nodes are very low.

This approach has gained broad support. Five major national and international guidelines now endorse sentinel node mapping as an appropriate alternative to full pelvic lymph node removal for uterine-confined endometrioid tumors.⁶ Studies have shown high detection rates, strong sensitivity, and no compromise in long-term survival compared with traditional lymph node dissection.⁷

One finding that underscores why sentinel mapping matters: in one study, patients who underwent sentinel node mapping had roughly three times the odds of having metastatic disease detected compared with those who had standard lymph node removal alone. The sentinel node was the only positive node in half of cases with metastatic spread, and about two-thirds of the metastases found were tiny deposits — micrometastases or isolated tumor cells — that would likely have been missed by routine pathology.⁸ In other words, sentinel mapping doesn’t just reduce the extent of surgery; it actually catches more disease, because the identified nodes receive more thorough pathological examination.

Why Lymphovascular Space Invasion Matters Now

The 2023 staging system explicitly incorporates lymphovascular space invasion (LVSI) — the presence of tumor cells within blood or lymph vessels in the uterine wall. Under the old system, LVSI was noted in pathology reports but didn’t directly change the stage. Now it does, because evidence has accumulated showing it’s a meaningful predictor of recurrence.

A study of nearly 1,800 patients with early-stage disease found that the extent of LVSI made a substantial difference. Patients with substantial LVSI had a five-year progression-free survival of about 82%, compared with roughly 90% for focal LVSI and 95% for no LVSI. Substantial LVSI was an independent predictor of worse outcomes, with about two and a half times the risk of progression compared to no LVSI. Interestingly, focal LVSI did not differ from no LVSI in a statistically meaningful way.⁹ The distinction between “a little” and “a lot” of LVSI turns out to matter quite a bit, and the new staging system reflects that.

Preoperative Imaging and Its Limits

Before surgery, imaging helps estimate how advanced the cancer is. MRI is the most commonly used modality for assessing how deeply the tumor has grown into the muscle wall of the uterus (myometrial invasion) and whether the cervix is involved. But MRI isn’t perfect. One study found an overall staging accuracy of about 74%, with depth of invasion underestimated more often than overestimated.¹⁰ MRI performs best at detecting deep myometrial invasion, where earlier work reported sensitivity and specificity around 88% and 85% respectively, and less well for superficial invasion or tumor confined to the lining.¹¹

Adding diffusion-weighted imaging to standard MRI sequences improves accuracy. Studies using high-field MRI with diffusion-weighted sequences showed strong correlation with final pathology for depth of invasion and outperformed contrast-enhanced MRI alone.¹²

Transvaginal ultrasound is a less expensive alternative and is more widely available. Pooled data suggest it detects deep myometrial invasion with about 82% sensitivity and 81% specificity.¹³ Subjective expert assessment by an experienced sonographer tends to outperform objective measurement-based methods, reaching about 79% sensitivity and roughly 76% overall accuracy.¹⁴ In practice, the choice between MRI and ultrasound often depends on local expertise and equipment availability.

For detecting distant spread, PET/CT scans play a growing role. A multicenter trial found that PET/CT had about 65% sensitivity and nearly 99% specificity for identifying distant metastasis in endometrial cancer, with a high positive predictive value around 86%.¹⁵ That high specificity means a positive finding on PET/CT is very likely to be real, which is valuable for deciding whether a patient has stage IV disease. PET/CT has also shown particular value in picking up hidden metastatic deposits that other imaging misses.¹⁶

Aggressive Histological Subtypes Behave Differently at Every Stage

Not all endometrial cancers are created equal, and staging has always struggled with this. The most common type, endometrioid carcinoma, tends to be caught early and has relatively favorable outcomes stage for stage. But papillary serous and clear cell variants are more aggressive. Even when apparently confined to early stages, these subtypes recur far more often. Patients with papillary serous or clear cell tumors that had minimal myometrial invasion had a five-year survival of about 56%, compared to 93% for endometrioid tumors with the same anatomy. Even tumors confined entirely to the endometrial lining showed a five-year survival of 60% for the aggressive subtypes versus 98% or higher for endometrioid cancers.¹⁷

These striking differences are one reason the 2023 system moves beyond anatomy. Under the old framework, a papillary serous tumor sitting on the surface of the uterine lining and an endometrioid tumor doing the same thing received the same stage, even though their recurrence risk differed by a factor of four or more. The updated system tries to capture that difference through both histological typing and molecular profiling.

Cervical Involvement and Stage II

Stage II applies when the cancer has grown from the uterine body into the cervical stroma. This is a relatively uncommon presentation, and the prognosis depends heavily on factors beyond the cervical invasion itself. In a study of stage II endometrioid adenocarcinomas, the strongest predictors of survival by multivariate analysis were age, lymphovascular space invasion, and the type of treatment received — not the depth of cervical invasion per se.¹⁸ This finding reinforces the broader principle that anatomy alone is an incomplete picture. Two stage II patients with identical cervical involvement can face very different odds depending on vessel invasion, grade, and now molecular subtype.

Biopsy Versus Final Pathology

A persistent challenge is that preoperative biopsies don’t always match the final surgical pathology. The histology from a biopsy taken before surgery can differ from the definitive specimen in a substantial fraction of cases — estimates range from about 15% to 25%, with some reports suggesting even higher discrepancy rates.¹⁹ The tumor might be upgraded to a higher grade, the histological type might change, or invasion might turn out to be deeper than the biopsy suggested. This means preoperative staging is always provisional. The final stage comes only after the pathologist has examined the entire surgical specimen, which is one reason definitive surgery remains central to the staging process.

Staging and Fertility Preservation

For younger patients who want to preserve the ability to have children, staging takes on a different urgency. Fertility-sparing treatment, typically involving hormonal therapy with progestins or a hormone-releasing intrauterine device, is considered only for a narrow window: grade 1 endometrioid tumors, stage IA, with no myometrial invasion and no high-risk features.²⁰ For grade 2 endometrioid tumors, the evidence is limited and decisions are made case by case.

The absence of myometrial invasion must be confirmed before this path is chosen, using either MRI or transvaginal ultrasound performed by a specialist. Because the consequences of underestimating invasion are serious — hormonal treatment cannot address cancer that has grown into the muscle wall — the accuracy limitations of preoperative imaging discussed earlier take on heightened importance in this specific context. These hormonal approaches have been shown to be feasible and safe in patients with early-stage disease and no significant invasion.²¹

How the Updated Staging System Affects Treatment Decisions

Restaging patients from the 2009 system to the 2023 system doesn’t just reclassify them on paper. It changes what treatment they’re recommended. When one group of researchers re-evaluated early-stage patients under the new system, they found a statistically significant decrease in the likelihood of receiving adjuvant treatment overall, with a relative risk of about 0.84 compared to the old system. At the same time, there was no significant increase in the risk of being assigned more aggressive treatment.²² In practical terms, the new system appears to be better at identifying patients who can safely skip post-surgical radiation or chemotherapy, without under-treating those who genuinely need it.

Among patients who had been candidates for vaginal brachytherapy under the old system, nearly half were reclassified to follow-up only under the new system. Some patients moved in the opposite direction, from radiation alone to combination chemotherapy and radiation. The overall effect is a redistribution: fewer patients treated when they don’t need it, and a more targeted escalation for those who do.

Global Access and the Molecular Staging Gap

The enthusiasm for molecular staging runs into a practical wall: not every hospital, and not every country, can perform the testing. Molecular classification requires specialized pathology infrastructure, and in resource-limited settings this is often unavailable. Critics have pointed out that incorporating molecular parameters essentially makes full staging impossible in many parts of the world where endometrial cancer is still common.²³ Limited availability of molecular profiling hinders universal adoption of the system.²⁴

The 2023 system handles this by making molecular classification “encouraged” rather than mandatory. A patient can still be staged without it — the anatomical framework remains intact as a standalone pathway. But this creates a two-tier reality: patients at well-resourced centers may benefit from stage adjustments that refine their treatment, while patients elsewhere are staged with less precision. Whether this gap widens existing disparities in cancer outcomes is an open question the field hasn’t fully reckoned with.

Adding to the concern, the new system includes variables whose definitions are still evolving, and some pathological features incorporated into staging are subject to considerable variation between pathologists.²⁵ Even with molecular tools available, two pathologists examining the same specimen might not always agree on the substage.

Emerging Technologies in Staging

Researchers are exploring whether artificial intelligence can improve preoperative staging by analyzing MRI scans more systematically than the human eye can. One deep learning pipeline designed to predict deep myometrial invasion from MRI achieved a balanced accuracy of about 70%, compared with about 77% for expert radiologists reading the same scans.²⁶ That gap may narrow as models are trained on larger datasets. Separate work using machine learning on MRI-derived radiomics features has shown promise in distinguishing early from advanced-stage disease and separating low-risk from high-risk tumors before surgery.²⁷

Liquid biopsy is another frontier. The idea is to detect circulating tumor cells or fragments of tumor DNA in a blood sample, which could provide staging-relevant information without surgery. Early work suggests these circulating biomarkers could enable non-invasive, real-time monitoring of how the cancer is evolving.²⁸ Neither AI-based imaging nor liquid biopsy is ready to replace surgical staging, but both represent areas where the staging process could become less invasive and more precise over the next decade.

Rare Presentations at Stage IV

Most endometrial cancers are diagnosed at stage I, when the tumor is still confined to the uterus. Stage IV, where cancer has spread to distant organs, is uncommon at initial diagnosis but not unheard of. In rare cases, the very first sign of endometrial cancer is a problem somewhere else entirely — a bone fracture, for instance. Case reports have documented endometrial cancer presenting initially as bone metastasis, even in the absence of any gynecological symptoms. When clinicians investigate bone tumors of unknown origin, endometrial biopsy probably isn’t the first thing on the list, but these cases illustrate why it belongs in the workup for women with metastatic adenocarcinoma without a clear source.²⁹ Interestingly, even in this scenario, survival outcomes appear relatively favorable compared with other advanced cancers presenting the same way.