CD138 and Recurrent Miscarriage: Does Doxycycline Help?

A UK randomized trial of 438 women compared doxycycline with placebo across 27 centers.

The primary outcome occurred in 51.6% versus 49.8%, raising questions about treatment and diagnostic criteria.

Does treating chronic endometritis with antibiotics improve the chances of a baby after recurrent miscarriage? A randomized trial found no clear improvement in birth outcomes when women with endometrial inflammation-related markers received doxycycline.

Researchers including teams from the Universities of Warwick and Oxford studied 438 women with recurrent miscarriage and positive biopsy findings. They found no clear difference in birth outcomes or miscarriage between antibiotic and placebo groups.

The study appeared in the October 2026 issue of Human Reproduction and challenges the assumption that an improved inflammation test necessarily translates into a better pregnancy outcome.

Why have antibiotics been used for chronic endometritis?

Chronic endometritis refers to persistent inflammation of the uterine lining. It often causes no obvious pain or fever and may be identified during investigations of infertility or recurrent miscarriage.

Earlier research suggested that it could affect embryo implantation and pregnancy maintenance. Some fertility clinics consequently prescribe antibiotics before another conception attempt when an endometrial biopsy shows inflammation-related findings.

Doxycycline is a commonly used antibiotic in this setting. But changing a biopsy result and reducing the risk of miscarriage are different outcomes. A clear improvement in the chance of birth had not been established.

A double-blind trial across 27 UK centers

The researchers recruited women aged at least 18 and under 42 who had experienced at least two consecutive early miscarriages.

Of those with the protein marker CD138 identified on endometrial biopsy, 438 entered the randomized trial. CD138 is used to help identify cells associated with inflammation in the uterine lining.

The researchers assigned 219 participants to doxycycline, 100 mg twice daily for 14 days, and 219 to a matching placebo without an active drug ingredient.

Neither participants nor treating clinicians knew which treatment had been allocated, reducing the influence of expectations on the results.

The primary outcome occurred in 51.6% versus 49.8%

The primary outcome was having a live birth or maintaining a pregnancy to at least 24 weeks after treatment.

It was reached by 113 of 219 women receiving doxycycline (51.6%) and 109 of 219 receiving placebo (49.8%).

The difference was approximately 1.8 percentage points. Statistical analysis did not provide clear evidence that antibiotics improved this outcome.

There was also no meaningful difference in miscarriage or time to pregnancy. Further analyses considering age, previous miscarriage history and the extent of endometrial findings did not identify a consistent benefit from doxycycline.

Biopsy findings also changed without antibiotics

Some participants underwent repeat biopsy to assess changes in the inflammation-related marker.

Results became negative in 33 of 104 women in the antibiotic group (31.7%) and 26 of 98 in the placebo group (26.5%). There was no clear difference between groups.

A negative result after antibiotics therefore cannot automatically be attributed to the drug. The researchers suggested that natural changes in the uterine lining across the menstrual cycle might affect the findings.

The endometrium thickens, matures and sheds in response to hormones each month. Its immune-cell distribution and tissue characteristics can change during that process.

Does a positive CD138 result always mean bacterial infection?

The study also raises questions about the definition of chronic endometritis.

Diagnosis commonly relies on identifying plasma cells in endometrial tissue, aided by CD138 staining. However, CD138 positivity does not necessarily establish inflammation caused by bacterial infection.

This trial used a broader definition that included other CD138-positive endometrial cells as well as typical plasma cells. The researchers suggested that such findings could reflect endometrial maturation as well as infectious inflammation.

Women receiving the same diagnostic label may therefore have findings with different underlying causes.

Treating the cause matters more than normalizing a test

The results do not show that antibiotics are unnecessary for every patient with chronic endometritis. They cannot simply be extended to confirmed bacterial infection or to patients diagnosed using different criteria.

Other limitations include stopping the trial at an interim-analysis stage and performing post-treatment biopsies within the same menstrual cycle.

Nevertheless, the findings support caution about repeatedly prescribing antibiotics solely on the basis of an inflammation-related test in recurrent miscarriage.

Possible contributors to recurrent miscarriage include uterine structural abnormalities, chromosome problems, antiphospholipid syndrome and thyroid dysfunction. A broad evaluation and treatments supported by outcome evidence matter more than focusing on a single result.

The researchers found no evidence in this trial that doxycycline improved birth outcomes, miscarriage, time to pregnancy or repeat biopsy findings.

The treatment goal is a sustained pregnancy and a baby—not simply making a laboratory or biopsy result negative.

Study source

J. Odendaal, N. Black, K. Fishwick, S. Quenby and colleagues, including the Universities of Warwick and Oxford: “A double-blind multi-centre randomised placebo-controlled trial of doxycycline in women with recurrent miscarriage and chronic endometritis defined by CD138 positive stromal cells.” Human Reproduction, 2026;41(10):1661–1671. Published online August 20, 2026. DOI: 10.1093/humrep/deag133. PMID: 42618515.

The illustration was generated using ChatGPT (OpenAI). Any people depicted are fictional.