Massive necrosis of cervical ectopic decidua presenting in labour


  • A. S. Gornall,

    Subspecialty Trainee (Fetomaternal Medicine), Corresponding author
    1. Departments of Obstetrics and Gynaecology, Leicester Royal Infirmary
      Correspondence: Mr A. S. Gornall, Department of Obstetrics and Gynaecology, Leicester Royal Infirmary Women's Hospital, Infirmary Square, Leicester LEI 5WW, UK.
    Search for more papers by this author
  • N. J. Naftalin,

    1. Departments of Obstetrics and Gynaecology, Leicester Royal Infirmary
    Search for more papers by this author
  • L. J. R. Brown,

    Consultant (Gynaecological Pathology)
    1. Department of Pathology, Leicester Royal Infirmary
    Search for more papers by this author
  • J. C. Konje

    Senior Lecturer/Consultant
    1. University of Leicester, Leicester Royal Infirmary
    Search for more papers by this author

Correspondence: Mr A. S. Gornall, Department of Obstetrics and Gynaecology, Leicester Royal Infirmary Women's Hospital, Infirmary Square, Leicester LEI 5WW, UK.

Case report

A 23 year old nulliparous woman presented with a history of spontaneous rupture of membranes unassociated with contractions at 38 weeks of gestation. She smoked 15 cigarettes per day and her last cervical smear two years previously was normal. The fetal head was engaged, clear amniotic fluid was observed draining from the vagina and the fetal heart rate was uncomplicated on cardiotocography. Her temperature was raised at 37.3°C.

Two weeks before admission she had presented to her general practitioner with an offensive vaginal discharge and was treated with antibiotics without examination or investigation. The discharge had persisted and was still present until rupture of the fetal membranes.

In view of the pyrexia she was offered induction of labour. A vaginal examination revealed a malodorous exophytic necrotic cervical mass. The lesion was hard and irregular associated with thickening of the parametrial tissues that did not extend to the pelvic side walls or the vagina. The cervical os could not be identified. A presumptive diagnosis of carcinoma of the cervix was made and multiple punch biopsies were taken for immediate histology by frozen section. She was then given intravenous cefuroxime and metronidazole. The histology report, obtained after two hours, was inconclusive because the tissue was too necrotic for diagnosis. Some inflammatory cells could be identified but no neoplastic cells were noted.

Because of suspected chorioamnionitis and with the possibility that the lesion was a cervical carcinoma, a caesarean section was performed under general anaesthesia. A midline sub-umbilical incision was made to allow complete exploration of the abdominal contents. The liveborn female infant had a widespread macular rash, and the amniotic fluid was malodorous consistent with chorioamnionitis; however, microbiological cultures were negative.

Exploration of the abdomen revealed discrete, tiny (approximately 0.5 × 0.5 cm) mobile para-aortic nodes. The liver surface was clear and rectal examination was unremarkable. Further examination of the cervix revealed a friable partially exophytic tumour measuring 8 cm in diameter which bled on contact. The overlying epithelium was necrotic centrally and inflamed at the edges of the lesion. The parametrial tissues were thought to be involved bilaterally. Three further cervical biopsies were taken from what appeared to be the tumour margin. Histology of the cervical biopsies (Fig. 1) showed extensive necrosis and inflammation with areas of necrotic decidual change. There were no neoplastic cells noted in any of the biopsy specimens. Four days after surgery a colposcopy was performed. The cervical lesion had regressed significantly and there were now only small areas of necrosis visible microscopically.

Figure 1.

Necrotic, plump spindle shaped decidualised stromal cells stand out against a background of karrhyohectic debris (× 250 original magnification).

The baby was admitted to the neonatal intensive care unit where she was treated with antibiotics for five days. Mother and baby were discharged home on the eighth day. At repeat colposcopy one month later there was no discharge, and the cervix had reverted to a completely normal appearance.


During pregnancy progesterone, and possibly other hormones, cause decidualisation of the endometrial stroma1. Ectopic decidual reactions, also known as deciduosis, can occur in the cervix, vagina,2,3 ovary, omentum,4 appendix,5 peritoneum4 and pelvic lymph nodes6. Deciduosis may be related to endometriosis, but published literature has not shown a consistent association, and the reaction regresses completely following withdrawal of hormones at delivery4

Cervical deciduosis was first described over 100 years ago7 and since then a few reports have been published either as a result of an unusual clinical appearance2,3,8 or as a discussion concerning the natural history of carcinoma of the cervix. Cervical deciduosis is more common as pregnancy progresses, which may explain the variation in the reported prevalence of 10–100%9–11. Most cases have been asymptomatic, displaying microscopic nodules or more diffuse alterations with occasional mild chronic inflammation. In four cases cervical deciduosis appeared as a large lesion initially thought to be cervical carcinoma2,3,8. All of these cases appeared antenatally, one in the first trimester, two in the second trimester and one in the third trimester. None was found during labour although one case was found just before labour commenced. In our case there was also clinical confusion with a carcinoma of the cervix.

The normal changes of the cervix during pregnancy include softening, cyanosis, hypertrophy, increased vascularity and hyperplasia of the endocervical glands, including Arias-Stella change12–15. Microglandular hyperplasia of the cervix may appear as a polypoid or nodular excrescence and has been reported in pregnancy13,16. Biopsy is necessary to provide the histological diagnosis for a cervical lesion, but if a lesion is discovered during labour rapid diagnosis may not be possible. Because of sampling, necrosis or artefact frozen section may pose diagnostic difficulties.

Vaginal delivery may be complicated by a friable cervical lesion that may bleed as the cervix dilates. There is also a theoretical concern that vaginal delivery may disseminate a cervical malignancy17. However, in three of the previously reported examples of cervical deciduosis diagnosed before labour, vaginal delivery occurred without problem. The fourth example was undiagnosed prior to labour but the lesion involved only the anterior lip of the cervix allowing uncomplicated vaginal delivery. In our case the initial biopsy was necrotic, and we could not exclude neoplasia. Caesarean section provided a means of avoiding the risk of both massive haemorrhage from a friable dilating cervix and dissemination of malignancy. The peri-operative biopsy subsequently indicated extensively necrotic decidua and the lesion regressed following delivery. If the exact diagnosis had been known prior to delivery, caesarean section may have been avoided in the light of the previously described cases.

In conclusion, a cervical lesion discovered during pregnancy or labour that on initial viewing may appear to be an advanced carcinoma of the cervix may represent a more benign process. Histological diagnosis is imperative for informed management of delivery.