What is the appropriate management for a female patient of reproductive age with endometrial stromal hyperplasia?

Management of Endometrial Stromal Hyperplasia in Reproductive-Age Women

Critical Clarification: Terminology and Diagnostic Approach

The term "stromal hyperplasia" is distinct from endometrial hyperplasia and requires immediate tissue diagnosis to clarify the pathology. True endometrial stromal hyperplasia is rare and differs fundamentally from the more common endometrial glandular hyperplasia (which affects glands, not stroma). 1

• Endometrial biopsy is mandatory to establish the correct diagnosis, as stromal proliferations can represent benign stromal hyperplasia, endometrial stromal nodules, or low-grade endometrial stromal sarcoma—entities with vastly different management implications 1
• Office endometrial biopsy may not accurately diagnose mesenchymal tumors or endometrial stromal neoplasms, so if initial sampling is non-diagnostic and clinical suspicion persists, proceed to hysteroscopy with directed biopsy or fractional D&C 1, 2

If Diagnosis Confirms Endometrial Glandular Hyperplasia (Most Likely Scenario)

For Hyperplasia Without Atypia

Levonorgestrel-releasing intrauterine system (LNG-IUS) is the preferred first-line treatment for reproductive-age women with non-atypical endometrial hyperplasia, offering higher regression rates, lower recurrence rates, and fewer adverse events compared to oral progestins 3

• Alternative: Oral progestins (medroxyprogesterone acetate or megestrol acetate) if LNG-IUS is contraindicated or declined 3, 4
• Monitor with endometrial biopsy every 6 months during treatment until two consecutive biopsies show no pathological changes 3
• Hysterectomy is not the preferred treatment for non-atypical hyperplasia in reproductive-age women 3

For Atypical Hyperplasia/Endometrial Intraepithelial Neoplasia (EIN)

Hysterectomy with bilateral salpingectomy is the definitive treatment for atypical hyperplasia, as it carries an 8-29% risk of progression to endometrial cancer and up to 40% risk of concurrent occult cancer 5, 6, 4, 7

• For women desiring fertility preservation: LNG-IUS is the preferred medical therapy, with endometrial pathologic evaluation every 3 months and treatment continuation until two consecutive biopsies show complete regression 3, 4
• After complete regression, women desiring pregnancy should pursue assisted reproductive technologies promptly, as recurrence risk is substantial 3
• If medical therapy fails or disease recurs: Total hysterectomy is recommended 3

Risk Factor Assessment and Prevention

Evaluate for underlying causes of unopposed estrogen exposure:

• Polycystic ovary syndrome (PCOS)—the most common cause in reproductive-age women with chronic anovulation 8, 4
• Obesity, which increases circulating estrogen levels 8, 4
• Lynch syndrome—requires genetic counseling and annual endometrial biopsy surveillance starting at age 30-35 years due to 30-60% lifetime endometrial cancer risk 5, 2

Preventive strategies include:

• Combined oral contraceptives or progestin-containing IUDs for women with chronic anovulation 4
• Weight reduction counseling for obese patients 8

Common Pitfalls to Avoid

• Never accept inadequate tissue sampling as reassuring in a symptomatic patient—office biopsy has a 10% false-negative rate 1, 2
• Do not delay tissue diagnosis by initiating empiric hormonal therapy without establishing the pathology first 2
• Do not confuse stromal hyperplasia with glandular hyperplasia—they are distinct entities requiring different management 1
• Do not perform lymphadenectomy for hyperplasia with or without atypia, as there is no indication for nodal assessment 3

Long-Term Follow-Up

After successful treatment, long-term surveillance is essential:

• Continue monitoring even after complete regression, as recurrence rates are significant 3
• Patient education regarding medication adherence is critical for improving regression rates and lowering recurrence 3
• Any recurrent abnormal bleeding requires immediate re-evaluation with repeat endometrial sampling 2, 4