Miscarriage Associated with Zika Virus Infection

Abstract

A 31-year-old woman who was 10 weeks pregnant contracted Zika virus infection in Suriname; this led to fetal loss. ZIKV was detected in the woman's blood for at least 21 days.

To the Editor: Data linking Zika virus (ZIKV) infection to fetal death have been reported in only a handful of cases to date.¹,² Here, we present a case of ZIKV infection in a woman who had a miscarriage. ZIKV was detected in the fetal tissue, and ZIKV viremia lasted for at least 21 days.

In January 2016, a 31-year-old woman who was 10 weeks pregnant visited an outpatient clinic in Rotterdam, the Netherlands, because of a 2-day history of headache, mild arthralgia in both wrists and the left knee, and a pruritic, macular rash. The symptoms had begun the day after her return from a 3.5-week trip to Suriname, which borders on northern Brazil. During her visit, she had not used malaria chemoprophylaxis or personal protective measures, such as insect repellents.

The patient’s medical history was uneventful, as was her pregnancy. When she was a child, she had received vaccines according to the regular vaccination program in Suriname, where she had lived until her early twenties. She was vaccinated against yellow fever at the age of 5 years.

A physical examination revealed normal vital-sign measurements and an absence of fever. A discrete macular rash was noted on the patient’s trunk and both arms. The joints were not visibly swollen or warm but mildly painful with movement. Blood and urine samples were obtained at regular intervals after the physical examination.

The patient’s symptoms resolved spontaneously after 6 days. On day 14 after the onset of symptoms, ultrasonography performed at a routine prenatal screening visit (estimated gestational age, 11 weeks and 4 days) showed no fetal heartbeat.

One week later, 21 days after the onset of ZIKV-associated disease, amniocentesis was performed, followed by dilation and curettage. Amniotic fluid and fetal and placental tissue were obtained for further laboratory analysis. (A detailed description of the methods is provided in the Supplementary Appendix, available with the full text of this letter at NEJM.org.)

The amniotic fluid and fetal and placental tissue tested positive for ZIKV on semiquantitative real-time reverse-transcriptase–polymerase-chain-reaction (RT-PCR) assays and cell-culture isolation; these results were confirmed by means of whole-genome sequencing (GenBank accession number, KU937936) (Fig. S1 through S4 in the Supplementary Appendix). The viral isolate (reference number, 011V-01621) can be obtained through the European Virus Archive Goes Global catalogue (www.european-virus-archive.com/virus/zika-virus-strain-suriname-2016).

Serum and urine samples obtained from the patient were positive for ZIKV on semiquantitative RT-PCR at multiple time points. A serum specimen was still positive for ZIKV on semiquantitative RT-PCR 21 days after the onset of clinical symptoms, although no ZIKV was detected in urine specimens after day 12. A more than quadrupling of neutralizing antibody titers against ZIKV was observed between days 2 and 12. On day 28, ZIKV RNA was no longer detected in serum specimens (see Table S1 in the Supplementary Appendix). Possible genetic disorders and other congenital infections were not identified (see Sections 8 and 9 in the Supplementary Appendix).

The findings on histopathological analysis of placental tissue specimens were consistent with intrauterine fetal death 1 week before curettage was performed. Histopathological and immunohistochemical investigation with the use of CD45 and CD3 antibodies to detect inflammation did not show evidence of increased inflammatory-cell infiltration. In situ hybridization with the use of ZIKV-specific probes, as compared with control probes, revealed that placental amniotic epithelium was positive for ZIKV RNA, whereas fetal chorion and trophoblasts and maternal decidua did not show evidence of ZIKV infection. In addition, positive staining of fetal mesenchymal cells with affinity for perichondrium was observed (Figure 1). Because of autolysis, no recognizable fetal brain tissue was present. Other fetal tissues, including those from the cartilage, kidneys, adrenal glands, gut, and eyes, were negative for ZIKV. These observations were confirmed on immunostaining of double-stranded RNA in fetal tissues (Fig. S5 in the Supplementary Appendix).

We found evidence of ZIKV infection in amniotic epithelial cells and in fetal mesenchymal cells with affinity for perichondrium. Our observation indicates that ZIKV replicates in pluripotent (amniotic stem) cells involved in early-stage embryo development. The observation of prolonged viremia until day 21 in the patient is in concordance with the findings of Driggers et al.³ and provides further data for consideration in the ongoing development of testing algorithms in pregnant women. These algorithms are currently based on the assumption that ZIKV viremia can be detected only up to 7 days.

Drs. van der Eijk, van Genderen, and Verdijk contributed equally to this letter.

This letter was published on July 27, 2016, at NEJM.org.

Supported by the Horizon 2020 research and innovation program of the European Union (grant agreement no. 643476) and by the European Virus Archive Goes Global project, which received a grant (653316) from the European Union Horizon 2020 Framework Program for Research and Innovation.

Disclosure forms provided by the authors are available with the full text of this letter at NEJM.org.