Endometrial Hyperplasia: With or Without Atypia

Endometrial hyperplasia means the lining of the uterus has genuinely overgrown. It is not a cancer. It is a diagnosis made on a tissue sample, not on a scan, and it comes in two forms that behave so differently they are best thought of as two separate conditions that share a name.

One is common, low-risk and usually reversed with a hormonal treatment. The other carries a substantial risk of a cancer being present already and is usually managed with surgery. The word that separates them is atypia.

This page is written for the woman holding a histology report. If you are instead holding a scan report that says your lining is thickened, that is a different and earlier stage of the process — our page on what a thickened endometrium on a scan actually means is the one you want first.

What hyperplasia actually is

The lining of the uterus is built up by oestrogen and stabilised, then shed, under the influence of progesterone. In a normal ovulatory cycle those two hormones alternate, and the lining is cleared out every month.

When oestrogen acts on the lining without enough progesterone to oppose it, month after month and year after year, the glands crowd together and the lining thickens beyond what it should. That is hyperplasia. It is a response to a hormonal imbalance, not a random event, which is why the treatment is largely about supplying the missing hormone and removing the source of the imbalance.

The commonest reasons for unopposed oestrogen are worth naming, because most of them are things that can be changed:

• A raised BMI. Fat tissue converts other hormones into oestrogen, so a higher body weight means a higher oestrogen level reaching the lining. After the menopause, when the ovaries have stopped, this becomes the main source of oestrogen in the body. It is the strongest modifiable risk factor there is.
• Infrequent or absent ovulation — PCOS being the commonest cause, and the years approaching menopause the second. If you do not ovulate, you do not produce progesterone, and the lining is never properly opposed. See our page on PCOS.
• Oestrogen-only hormone therapy in a woman who still has a uterus. This is why HRT for a woman with a uterus always includes a progestogen.
• Tamoxifen, used after breast cancer. It blocks oestrogen in the breast but stimulates the uterine lining, and the risk rises with age above 50.
• Oestrogen-secreting ovarian tumours — uncommon, but a reason the ovaries are looked at too.
• Diabetes and long-term immunosuppression are both associated.

The two types, and why the distinction is everything

The World Health Organization classification divides endometrial hyperplasia into exactly two categories, based on whether the individual cells look abnormal down the microscope:

Hyperplasia without atypia

The glands are crowded and overgrown, but the cells themselves look normal. This is the commoner diagnosis by a wide margin.

The key number: the risk of progressing to endometrial cancer is under 5% over 20 years. That is a low figure over a very long window, and it is the reason this version is treated medically rather than surgically. A substantial proportion even regress on their own — historical series put spontaneous regression at around 75 to 80% — though "it might go away by itself" is not a treatment plan, because treatment works considerably better and faster.

Atypical hyperplasia

Here the cells themselves are abnormal. You may see it written as atypical hyperplasia, or as endometrioid intraepithelial neoplasia (EIN) — different pathology schools use different terminology for what is, in practice, the same finding. If your report uses the word EIN, read it as atypical hyperplasia.

This is a pre-cancerous condition in the meaningful sense of the term. The cumulative risk of developing endometrial cancer is:

• 8% at four years
• 12.4% at nine years
• 27.5% at nineteen years

How the diagnosis is made

Hyperplasia cannot be diagnosed on an ultrasound. A scan can raise the question — a lining above 4 mm after the menopause needs explaining, and above about 7 mm in a woman with PCOS makes hyperplasia more likely — but only tissue can answer it.

The sample is usually taken in one of two ways: an outpatient endometrial biopsy, using a fine flexible tube passed through the cervix; or hysteroscopy with sampling, in which a thin camera is passed into the cavity so the lining can be seen directly and targeted. Hysteroscopy is more accurate, particularly for detecting a focal abnormality such as a small cancer, which is exactly why it is preferred when atypia has been found.

Both of these are hospital or day-care procedures rather than clinic-room ones. At this practice, the consultation, the interpretation of your report and the treatment that follows happen at the clinic; hysteroscopy, sampling under anaesthetic and any surgery are arranged at hospital, and scans are arranged by referral.

Treating hyperplasia without atypia

The aim is to supply the progesterone the lining has been missing, and to keep supplying it until repeat sampling shows the lining has returned to normal.

The hormonal coil is first-line, and it is not close

The levonorgestrel intrauterine system — the Mirena — delivers progestogen directly onto the lining rather than through the whole body. Head to head against progestogen tablets, it produces better regression at every point measured, and the gap widens over time. Expressed as odds ratios, the coil outperformed tablets by roughly:

• 2.3 times at three months
• 3.2 times at six months
• 5.7 times at twelve months
• 7.5 times at twenty-four months

It also has fewer systemic side effects than tablets, because far less hormone reaches the rest of the body, and it removes the compliance problem entirely — a device cannot be forgotten. For a woman who also has heavy periods, it treats both at once. Our full guide to the Mirena covers what it is like to live with.

Fitting a Mirena is done here at the clinic as an outpatient procedure, with no admission and no general anaesthetic. That is worth knowing, because the treatment for this condition is often assumed to require a hospital visit and it usually does not.

Tablets, when a coil is not possible

Continuous oral progestogens — medroxyprogesterone acetate or megestrol acetate — are the alternative and do work. Two points matter: continuous dosing outperforms cyclical dosing, so a tablet taken for ten days a month is the weakest of the options; and women on tablets need closer follow-up than women with a coil.

How long, and then what

Treatment runs for a minimum of six months. In practice, a coil is usually left in place for longer — up to five years — because keeping it in substantially reduces the chance of relapse, and because there is no reason to remove something that is working.

Follow-up is by repeat endometrial sampling, not by scan and not by symptoms. The standard is a biopsy at least every six months, and two consecutive negative biopsies before you can be discharged from surveillance. Women at higher risk — a BMI of 35 or above, or those on tablets rather than a coil — continue with six-monthly sampling and then annual sampling long term.

Treating atypical hyperplasia

Here the default flips. Because of the 30 to 43% chance that a cancer is already present, and the substantial risk of one developing, total hysterectomy is the standard recommendation for a woman who has completed her family.

Two specifics are worth knowing, because both are occasionally got wrong:

• The cervix must come out too. A subtotal or supracervical hysterectomy is not appropriate for this diagnosis.
• Endometrial ablation is not a treatment for it. Burning or destroying the lining leaves tissue behind that cannot then be sampled, which is precisely the wrong thing to do when the question is whether a cancer is hiding.

In a woman past the menopause, removal of the tubes and ovaries is generally recommended alongside. Before the menopause it is an individual discussion. Whether a gynaecologic oncologist should be involved depends on the individual picture and on what is available; it is a reasonable thing to ask about.

If you still want to have children

This is where atypical hyperplasia becomes a genuinely difficult conversation rather than a protocol, and it can be handled well.

Fertility-sparing management is possible. The coil is first-line here too, sometimes combined with oral progestogens — one study found 85% regression with the combination against 55% for the coil alone. Overall, the initial response rate to progestogen treatment in atypical hyperplasia is around 86%, with complete resolution in about 66%.

The honest figures for what follows, from pooled data:

• 85.6% of women achieve regression of the disease
• 26.3% achieve a live birth
• 26% relapse

Surveillance during this is intensive: repeat sampling every three months until two consecutive negative biopsies, then continued sampling every three to six months for up to two years. And the recommendation at the end is unambiguous: hysterectomy once childbearing is complete, because the relapse rate does not go away.

This route is a considered decision made with full information, not a default and not a way of avoiding surgery. It is entirely reasonable to choose, and entirely reasonable to want a second opinion before choosing it.

The part you can actually influence

Treatment reverses the lining. It does not, on its own, remove the reason the lining overgrew in the first place. That is where the recurrence risk lives, and it is where you have more leverage than anywhere else in this condition.

Weight is the biggest single lever. In one study of women with atypical hyperplasia on a coil, those who lost more than 10% of their body weight were nearly four times more likely to respond to treatment. That is a treatment effect from weight loss comparable to a drug, and it is rarely presented that way.

Restoring ovulation matters if you are not ovulating. In PCOS, anything that re-establishes regular cycles — or supplies the progestogen that ovulation would have — protects the lining.

Hormone therapy can be reviewed. Oestrogen-only HRT in a woman with a uterus should be changed. Tamoxifen is a more difficult conversation, because the reason it is being taken usually outweighs the endometrial risk, but it should be an explicit discussion with the team treating the breast cancer rather than a unilateral decision.

Blood sugar control, where diabetes is present, belongs on the same list.

What to ask at your appointment

Bring the report itself, not a summary of it. Then five questions:

• "With atypia, or without?" Everything else follows from this.
• "How was the sample taken — outpatient biopsy or hysteroscopy?" If atypia was found on a blind outpatient sample, whether the cavity should be looked at directly before deciding is a fair question.
• "What is the plan for repeat sampling, and when?" A treatment plan without a follow-up biopsy date is incomplete.
• "What caused this in my case?" If nobody can name a reason, the reason has not been looked for.
• "If I still want children, what are my options?" Ask this early, before a plan is made around an assumption.

What happens at the clinic

Consultation, going through your histology and scan reports, planning treatment, and fitting a hormonal coil where that is the treatment, are all done here at the clinic — the coil is an outpatient procedure with no admission and no general anaesthetic. Blood tests are done here too.

Hysteroscopy, endometrial sampling under anaesthetic and any surgery, including hysterectomy, are arranged at hospital. Ultrasound scans are arranged by referral, since we do not run a scanner at the clinic. If you are coming for a second opinion on a report or a recommendation made elsewhere, that is welcome; bring the original report and any previous samples' results.