Maternal early primary CMV infection with negative or equivocal CMV IgM.

BACKGROUND

Accurate diagnosis of primary cytomegalovirus infection (CMV PI) during pregnancy is critical for prognosis and clinical management, yet CMV-IgM may be transient or absent in the early phase of infection, leading to underdiagnosis. We aimed to determine the frequency and characteristics of negative or equivocal CMV-IgM results observed in confirmed recent CMV PI and to assess whether this diagnostic pitfall could be reduced.

METHODS

We retrospectively analysed 1,090 cases of CMV PI occurring within 2 months of infection onset in pregnant women referred to the French National Reference Laboratory for Viral Perinatal Infections between 2013 and 2025. CMV-IgG, CMV-IgM, and CMV-IgG avidity were assessed using the LIAISON® XL platform, with additional avidity testing on VIDAS®. Recent infection was defined by IgG seroconversion and/or low IgG avidity (< 30% on VIDAS). We analysed the frequency, timing, and characteristics of negative or equivocal CMV-IgM results and evaluated the potential impact of pragmatic interpretation criteria.

RESULTS

Among 1,090 cases of early CMV PI, 2.6% (28/1,090) had negative or equivocal CMV-IgM results, including 1.7% (18/1,090) negative and 0.9% (10/1,090) equivocal results. CMV-IgM negativity was observed both within the first month [1.4% (9/664)] and between 1 and 2 months [2.1% (9/426)] after the onset of infection. Among IgM-negative patients, 39% (7/18) had low CMV-IgG levels on LIAISON® XL (< 28 U/mL), consistent with very early seroconversion. In contrast, 61% (11/18) had CMV-IgG levels above twice the positivity threshold (> 28 U/mL), mimicking past immunity and representing a true diagnostic blind spot, accounting for 1.0% (11/1,090) of all infections. Considering CMV-IgM values between 15 and 18 U/mL as low-reactive results requiring further assessment, together with repeat serology when CMV-IgG levels were below twice the positivity threshold, would have reclassified 14 of 18 IgM-negative cases and reduced missed diagnoses from 1.7% (18/1,090) to 0.4% (4/1,090).

CONCLUSION

Two complementary measures may improve detection of CMV-IgM-negative early infections while maintaining a manageable laboratory workload: repeat serology when CMV-IgG levels are very low, and further assessment of low-reactive CMV-IgM values between 15 and 18 U/mL. Nevertheless, a small residual diagnostic gap remains, highlighting the inherent limitations of CMV-IgM-based screening for recent CMV infection during pregnancy.